id
stringlengths 9
16
| submitter
stringlengths 3
64
⌀ | authors
stringlengths 5
6.63k
| title
stringlengths 7
245
| comments
stringlengths 1
482
⌀ | journal-ref
stringlengths 4
382
⌀ | doi
stringlengths 9
151
⌀ | report-no
stringclasses 984
values | categories
stringlengths 5
108
| license
stringclasses 9
values | abstract
stringlengths 83
3.41k
| versions
listlengths 1
20
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-11 00:00:00
| authors_parsed
listlengths 1
427
| prompt
stringlengths 166
3.49k
| label
stringclasses 2
values | prob
float64 0.5
0.98
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2405.03420 | Marcin Denkowski | Emil Benedykciuk and Marcin Denkowski and Grzegorz W\'ojcik | Implantable Adaptive Cells: A Novel Enhancement for Pre-Trained U-Nets
in Medical Image Segmentation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper introduces a novel approach to enhance the performance of
pre-trained neural networks in medical image segmentation using gradient-based
Neural Architecture Search (NAS) methods. We present the concept of Implantable
Adaptive Cell (IAC), small modules identified through Partially-Connected DARTS
based approach, designed to be injected into the skip connections of an
existing and already trained U-shaped model. Unlike traditional NAS methods,
our approach refines existing architectures without full retraining.
Experiments on four medical datasets with MRI and CT images show consistent
accuracy improvements on various U-Net configurations, with segmentation
accuracy gain by approximately 5 percentage points across all validation
datasets, with improvements reaching up to 11\%pt in the best-performing cases.
The findings of this study not only offer a cost-effective alternative to the
complete overhaul of complex models for performance upgrades but also indicate
the potential applicability of our method to other architectures and problem
domains.
| [
{
"version": "v1",
"created": "Mon, 6 May 2024 12:40:15 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 12:52:29 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Benedykciuk",
"Emil",
""
],
[
"Denkowski",
"Marcin",
""
],
[
"Wójcik",
"Grzegorz",
""
]
]
| TITLE: Implantable Adaptive Cells: A Novel Enhancement for Pre-Trained U-Nets
in Medical Image Segmentation
ABSTRACT: This paper introduces a novel approach to enhance the performance of
pre-trained neural networks in medical image segmentation using gradient-based
Neural Architecture Search (NAS) methods. We present the concept of Implantable
Adaptive Cell (IAC), small modules identified through Partially-Connected DARTS
based approach, designed to be injected into the skip connections of an
existing and already trained U-shaped model. Unlike traditional NAS methods,
our approach refines existing architectures without full retraining.
Experiments on four medical datasets with MRI and CT images show consistent
accuracy improvements on various U-Net configurations, with segmentation
accuracy gain by approximately 5 percentage points across all validation
datasets, with improvements reaching up to 11\%pt in the best-performing cases.
The findings of this study not only offer a cost-effective alternative to the
complete overhaul of complex models for performance upgrades but also indicate
the potential applicability of our method to other architectures and problem
domains.
| no_new_dataset | 0.945751 |
2405.04902 | Zhihan Ju | Zhihan Ju, Wanting Zhou, Longteng Kong, Yu Chen, Yi Li, Zhenan Sun,
Caifeng Shan | HAGAN: Hybrid Augmented Generative Adversarial Network for Medical Image
Synthesis | null | Machine Intelligence Research 2024 | 10.1007/s11633-024-1528-y | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Medical Image Synthesis (MIS) plays an important role in the intelligent
medical field, which greatly saves the economic and time costs of medical
diagnosis. However, due to the complexity of medical images and similar
characteristics of different tissue cells, existing methods face great
challenges in meeting their biological consistency. To this end, we propose the
Hybrid Augmented Generative Adversarial Network (HAGAN) to maintain the
authenticity of structural texture and tissue cells. HAGAN contains Attention
Mixed (AttnMix) Generator, Hierarchical Discriminator and Reverse Skip
Connection between Discriminator and Generator. The AttnMix consistency
differentiable regularization encourages the perception in structural and
textural variations between real and fake images, which improves the
pathological integrity of synthetic images and the accuracy of features in
local areas. The Hierarchical Discriminator introduces pixel-by-pixel
discriminant feedback to generator for enhancing the saliency and discriminance
of global and local details simultaneously. The Reverse Skip Connection further
improves the accuracy for fine details by fusing real and synthetic
distribution features. Our experimental evaluations on three datasets of
different scales, i.e., COVID-CT, ACDC and BraTS2018, demonstrate that HAGAN
outperforms the existing methods and achieves state-of-the-art performance in
both high-resolution and low-resolution.
| [
{
"version": "v1",
"created": "Wed, 8 May 2024 09:13:42 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Ju",
"Zhihan",
""
],
[
"Zhou",
"Wanting",
""
],
[
"Kong",
"Longteng",
""
],
[
"Chen",
"Yu",
""
],
[
"Li",
"Yi",
""
],
[
"Sun",
"Zhenan",
""
],
[
"Shan",
"Caifeng",
""
]
]
| TITLE: HAGAN: Hybrid Augmented Generative Adversarial Network for Medical Image
Synthesis
ABSTRACT: Medical Image Synthesis (MIS) plays an important role in the intelligent
medical field, which greatly saves the economic and time costs of medical
diagnosis. However, due to the complexity of medical images and similar
characteristics of different tissue cells, existing methods face great
challenges in meeting their biological consistency. To this end, we propose the
Hybrid Augmented Generative Adversarial Network (HAGAN) to maintain the
authenticity of structural texture and tissue cells. HAGAN contains Attention
Mixed (AttnMix) Generator, Hierarchical Discriminator and Reverse Skip
Connection between Discriminator and Generator. The AttnMix consistency
differentiable regularization encourages the perception in structural and
textural variations between real and fake images, which improves the
pathological integrity of synthetic images and the accuracy of features in
local areas. The Hierarchical Discriminator introduces pixel-by-pixel
discriminant feedback to generator for enhancing the saliency and discriminance
of global and local details simultaneously. The Reverse Skip Connection further
improves the accuracy for fine details by fusing real and synthetic
distribution features. Our experimental evaluations on three datasets of
different scales, i.e., COVID-CT, ACDC and BraTS2018, demonstrate that HAGAN
outperforms the existing methods and achieves state-of-the-art performance in
both high-resolution and low-resolution.
| no_new_dataset | 0.953966 |
2405.07155 | Hu Wang | Hu Wang, Salma Hassan, Yuyuan Liu, Congbo Ma, Yuanhong Chen, Yutong
Xie, Mostafa Salem, Yu Tian, Jodie Avery, Louise Hull, Ian Reid, Mohammad
Yaqub, Gustavo Carneiro | Meta-Learned Modality-Weighted Knowledge Distillation for Robust
Multi-Modal Learning with Missing Data | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In multi-modal learning, some modalities are more influential than others,
and their absence can have a significant impact on classification/segmentation
accuracy. Addressing this challenge, we propose a novel approach called
Meta-learned Modality-weighted Knowledge Distillation (MetaKD), which enables
multi-modal models to maintain high accuracy even when key modalities are
missing. MetaKD adaptively estimates the importance weight of each modality
through a meta-learning process. These learned importance weights guide a
pairwise modality-weighted knowledge distillation process, allowing
high-importance modalities to transfer knowledge to lower-importance ones,
resulting in robust performance despite missing inputs. Unlike previous methods
in the field, which are often task-specific and require significant
modifications, our approach is designed to work in multiple tasks (e.g.,
segmentation and classification) with minimal adaptation. Experimental results
on five prevalent datasets, including three Brain Tumor Segmentation datasets
(BraTS2018, BraTS2019 and BraTS2020), the Alzheimer's Disease Neuroimaging
Initiative (ADNI) classification dataset and the Audiovision-MNIST
classification dataset, demonstrate the proposed model is able to outperform
the compared models by a large margin. The code is available at
https://github.com/billhhh/MetaKD.
| [
{
"version": "v1",
"created": "Sun, 12 May 2024 04:18:10 GMT"
},
{
"version": "v2",
"created": "Tue, 12 Nov 2024 16:39:29 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 08:51:28 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"Hu",
""
],
[
"Hassan",
"Salma",
""
],
[
"Liu",
"Yuyuan",
""
],
[
"Ma",
"Congbo",
""
],
[
"Chen",
"Yuanhong",
""
],
[
"Xie",
"Yutong",
""
],
[
"Salem",
"Mostafa",
""
],
[
"Tian",
"Yu",
""
],
[
"Avery",
"Jodie",
""
],
[
"Hull",
"Louise",
""
],
[
"Reid",
"Ian",
""
],
[
"Yaqub",
"Mohammad",
""
],
[
"Carneiro",
"Gustavo",
""
]
]
| TITLE: Meta-Learned Modality-Weighted Knowledge Distillation for Robust
Multi-Modal Learning with Missing Data
ABSTRACT: In multi-modal learning, some modalities are more influential than others,
and their absence can have a significant impact on classification/segmentation
accuracy. Addressing this challenge, we propose a novel approach called
Meta-learned Modality-weighted Knowledge Distillation (MetaKD), which enables
multi-modal models to maintain high accuracy even when key modalities are
missing. MetaKD adaptively estimates the importance weight of each modality
through a meta-learning process. These learned importance weights guide a
pairwise modality-weighted knowledge distillation process, allowing
high-importance modalities to transfer knowledge to lower-importance ones,
resulting in robust performance despite missing inputs. Unlike previous methods
in the field, which are often task-specific and require significant
modifications, our approach is designed to work in multiple tasks (e.g.,
segmentation and classification) with minimal adaptation. Experimental results
on five prevalent datasets, including three Brain Tumor Segmentation datasets
(BraTS2018, BraTS2019 and BraTS2020), the Alzheimer's Disease Neuroimaging
Initiative (ADNI) classification dataset and the Audiovision-MNIST
classification dataset, demonstrate the proposed model is able to outperform
the compared models by a large margin. The code is available at
https://github.com/billhhh/MetaKD.
| no_new_dataset | 0.955026 |
2405.12833 | Ruizhe Li | Xinyi Wang, Grazziela Figueredo, Ruizhe Li, Wei Emma Zhang, Weitong
Chen, Xin Chen | A Survey of Deep Learning-based Radiology Report Generation Using
Multimodal Data | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatic radiology report generation can alleviate the workload for
physicians and minimize regional disparities in medical resources, therefore
becoming an important topic in the medical image analysis field. It is a
challenging task, as the computational model needs to mimic physicians to
obtain information from multi-modal input data (i.e., medical images, clinical
information, medical knowledge, etc.), and produce comprehensive and accurate
reports. Recently, numerous works have emerged to address this issue using
deep-learning-based methods, such as transformers, contrastive learning, and
knowledge-base construction. This survey summarizes the key techniques
developed in the most recent works and proposes a general workflow for
deep-learning-based report generation with five main components, including
multi-modality data acquisition, data preparation, feature learning, feature
fusion and interaction, and report generation. The state-of-the-art methods for
each of these components are highlighted. Additionally, we summarize the latest
developments in large model-based methods and model explainability, along with
public datasets, evaluation methods, current challenges, and future directions
in this field. We have also conducted a quantitative comparison between
different methods in the same experimental setting. This is the most up-to-date
survey that focuses on multi-modality inputs and data fusion for radiology
report generation. The aim is to provide comprehensive and rich information for
researchers interested in automatic clinical report generation and medical
image analysis, especially when using multimodal inputs, and to assist them in
developing new algorithms to advance the field.
| [
{
"version": "v1",
"created": "Tue, 21 May 2024 14:37:35 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 17:18:49 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"Xinyi",
""
],
[
"Figueredo",
"Grazziela",
""
],
[
"Li",
"Ruizhe",
""
],
[
"Zhang",
"Wei Emma",
""
],
[
"Chen",
"Weitong",
""
],
[
"Chen",
"Xin",
""
]
]
| TITLE: A Survey of Deep Learning-based Radiology Report Generation Using
Multimodal Data
ABSTRACT: Automatic radiology report generation can alleviate the workload for
physicians and minimize regional disparities in medical resources, therefore
becoming an important topic in the medical image analysis field. It is a
challenging task, as the computational model needs to mimic physicians to
obtain information from multi-modal input data (i.e., medical images, clinical
information, medical knowledge, etc.), and produce comprehensive and accurate
reports. Recently, numerous works have emerged to address this issue using
deep-learning-based methods, such as transformers, contrastive learning, and
knowledge-base construction. This survey summarizes the key techniques
developed in the most recent works and proposes a general workflow for
deep-learning-based report generation with five main components, including
multi-modality data acquisition, data preparation, feature learning, feature
fusion and interaction, and report generation. The state-of-the-art methods for
each of these components are highlighted. Additionally, we summarize the latest
developments in large model-based methods and model explainability, along with
public datasets, evaluation methods, current challenges, and future directions
in this field. We have also conducted a quantitative comparison between
different methods in the same experimental setting. This is the most up-to-date
survey that focuses on multi-modality inputs and data fusion for radiology
report generation. The aim is to provide comprehensive and rich information for
researchers interested in automatic clinical report generation and medical
image analysis, especially when using multimodal inputs, and to assist them in
developing new algorithms to advance the field.
| no_new_dataset | 0.947381 |
2405.14736 | Xinyi Shang | Xinyi Shang, Peng Sun, Tao Lin | GIFT: Unlocking Full Potential of Labels in Distilled Dataset at
Near-zero Cost | https://github.com/LINs-lab/GIFT | ICLR 2025 | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in dataset distillation have demonstrated the significant
benefits of employing soft labels generated by pre-trained teacher models. In
this paper, we introduce a novel perspective by emphasizing the full
utilization of labels. We first conduct a comprehensive comparison of various
loss functions for soft label utilization in dataset distillation, revealing
that the model trained on the synthetic dataset exhibits high sensitivity to
the choice of loss function for soft label utilization. This finding highlights
the necessity of a universal loss function for training models on synthetic
datasets. Building on these insights, we introduce an extremely simple yet
surprisingly effective plug-and-play approach, GIFT, which encompasses soft
label refinement and a cosine similarity-based loss function to efficiently
leverage full label information. Extensive experiments indicate that GIFT
consistently enhances state-of-the-art dataset distillation methods across
various dataset scales, without incurring additional computational costs.
Importantly, GIFT significantly enhances cross-optimizer generalization, an
area previously overlooked. For instance, on ImageNet-1K with IPC = 10, GIFT
enhances the state-of-the-art method RDED by 30.8% in cross-optimizer
generalization. Our code is available at https://github.com/LINs-lab/GIFT.
| [
{
"version": "v1",
"created": "Thu, 23 May 2024 16:02:30 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 09:52:43 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Shang",
"Xinyi",
""
],
[
"Sun",
"Peng",
""
],
[
"Lin",
"Tao",
""
]
]
| TITLE: GIFT: Unlocking Full Potential of Labels in Distilled Dataset at
Near-zero Cost
ABSTRACT: Recent advancements in dataset distillation have demonstrated the significant
benefits of employing soft labels generated by pre-trained teacher models. In
this paper, we introduce a novel perspective by emphasizing the full
utilization of labels. We first conduct a comprehensive comparison of various
loss functions for soft label utilization in dataset distillation, revealing
that the model trained on the synthetic dataset exhibits high sensitivity to
the choice of loss function for soft label utilization. This finding highlights
the necessity of a universal loss function for training models on synthetic
datasets. Building on these insights, we introduce an extremely simple yet
surprisingly effective plug-and-play approach, GIFT, which encompasses soft
label refinement and a cosine similarity-based loss function to efficiently
leverage full label information. Extensive experiments indicate that GIFT
consistently enhances state-of-the-art dataset distillation methods across
various dataset scales, without incurring additional computational costs.
Importantly, GIFT significantly enhances cross-optimizer generalization, an
area previously overlooked. For instance, on ImageNet-1K with IPC = 10, GIFT
enhances the state-of-the-art method RDED by 30.8% in cross-optimizer
generalization. Our code is available at https://github.com/LINs-lab/GIFT.
| no_new_dataset | 0.941169 |
2406.00799 | Sahar Abdelnabi | Sahar Abdelnabi, Aideen Fay, Giovanni Cherubin, Ahmed Salem, Mario
Fritz, Andrew Paverd | Get my drift? Catching LLM Task Drift with Activation Deltas | SaTML 2025 | null | null | null | cs.CR cs.CL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | LLMs are commonly used in retrieval-augmented applications to execute user
instructions based on data from external sources. For example, modern search
engines use LLMs to answer queries based on relevant search results; email
plugins summarize emails by processing their content through an LLM. However,
the potentially untrusted provenance of these data sources can lead to prompt
injection attacks, where the LLM is manipulated by natural language
instructions embedded in the external data, causing it to deviate from the
user's original instruction(s). We define this deviation as task drift. Task
drift is a significant concern as it allows attackers to exfiltrate data or
influence the LLM's output for other users. We study LLM activations as a
solution to detect task drift, showing that activation deltas - the difference
in activations before and after processing external data - are strongly
correlated with this phenomenon. Through two probing methods, we demonstrate
that a simple linear classifier can detect drift with near-perfect ROC AUC on
an out-of-distribution test set. We evaluate these methods by making minimal
assumptions about how users' tasks, system prompts, and attacks can be phrased.
We observe that this approach generalizes surprisingly well to unseen task
domains, such as prompt injections, jailbreaks, and malicious instructions,
without being trained on any of these attacks. Interestingly, the fact that
this solution does not require any modifications to the LLM (e.g.,
fine-tuning), as well as its compatibility with existing meta-prompting
solutions, makes it cost-efficient and easy to deploy. To encourage further
research on activation-based task inspection, decoding, and interpretability,
we release our large-scale TaskTracker toolkit, featuring a dataset of over
500K instances, representations from six SoTA language models, and a suite of
inspection tools.
| [
{
"version": "v1",
"created": "Sun, 2 Jun 2024 16:53:21 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Jun 2024 15:39:56 GMT"
},
{
"version": "v3",
"created": "Thu, 20 Jun 2024 13:33:08 GMT"
},
{
"version": "v4",
"created": "Fri, 19 Jul 2024 13:07:25 GMT"
},
{
"version": "v5",
"created": "Sun, 3 Nov 2024 14:52:35 GMT"
},
{
"version": "v6",
"created": "Thu, 6 Mar 2025 17:43:10 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Abdelnabi",
"Sahar",
""
],
[
"Fay",
"Aideen",
""
],
[
"Cherubin",
"Giovanni",
""
],
[
"Salem",
"Ahmed",
""
],
[
"Fritz",
"Mario",
""
],
[
"Paverd",
"Andrew",
""
]
]
| TITLE: Get my drift? Catching LLM Task Drift with Activation Deltas
ABSTRACT: LLMs are commonly used in retrieval-augmented applications to execute user
instructions based on data from external sources. For example, modern search
engines use LLMs to answer queries based on relevant search results; email
plugins summarize emails by processing their content through an LLM. However,
the potentially untrusted provenance of these data sources can lead to prompt
injection attacks, where the LLM is manipulated by natural language
instructions embedded in the external data, causing it to deviate from the
user's original instruction(s). We define this deviation as task drift. Task
drift is a significant concern as it allows attackers to exfiltrate data or
influence the LLM's output for other users. We study LLM activations as a
solution to detect task drift, showing that activation deltas - the difference
in activations before and after processing external data - are strongly
correlated with this phenomenon. Through two probing methods, we demonstrate
that a simple linear classifier can detect drift with near-perfect ROC AUC on
an out-of-distribution test set. We evaluate these methods by making minimal
assumptions about how users' tasks, system prompts, and attacks can be phrased.
We observe that this approach generalizes surprisingly well to unseen task
domains, such as prompt injections, jailbreaks, and malicious instructions,
without being trained on any of these attacks. Interestingly, the fact that
this solution does not require any modifications to the LLM (e.g.,
fine-tuning), as well as its compatibility with existing meta-prompting
solutions, makes it cost-efficient and easy to deploy. To encourage further
research on activation-based task inspection, decoding, and interpretability,
we release our large-scale TaskTracker toolkit, featuring a dataset of over
500K instances, representations from six SoTA language models, and a suite of
inspection tools.
| no_new_dataset | 0.924891 |
2406.01145 | Guangyi Liu | Guangyi Liu, Yongqi Zhang, Yong Li, Quanming Yao | Dual Reasoning: A GNN-LLM Collaborative Framework for Knowledge Graph
Question Answering | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) excel at intuitive, implicit reasoning. Guiding
LLMs to construct thought chains can enhance their deliberate reasoning
abilities, but also faces challenges such as hallucination. Knowledge Graphs
(KGs) can provide explicit structured knowledge for LLMs to alleviate these
issues. However, existing KG-enhanced methods often overlook explicit graph
learning, making it challenging to efficiently provide precise reasoning chains
for LLMs. Following dual-process theory, we propose Dual-Reasoning (DualR), a
novel framework that integrates an external system based on Graph Neural
Network (GNN) for explicit reasoning on KGs, complementing the implicit
reasoning of LLMs through externalized reasoning chains. DualR designs an
LLM-empowered GNN module for explicit learning on KGs, efficiently extracting
high-quality reasoning chains. These reasoning chains are then refined to a
knowledge-enhanced multiple-choice prompt, guiding a frozen LLM to reason
thoughtfully for final answer determination. Extensive experiments on three
benchmark KGQA datasets demonstrate that DualR achieves state-of-the-art
performance while maintaining high efficiency and interpretability.
| [
{
"version": "v1",
"created": "Mon, 3 Jun 2024 09:38:28 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 06:49:04 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Liu",
"Guangyi",
""
],
[
"Zhang",
"Yongqi",
""
],
[
"Li",
"Yong",
""
],
[
"Yao",
"Quanming",
""
]
]
| TITLE: Dual Reasoning: A GNN-LLM Collaborative Framework for Knowledge Graph
Question Answering
ABSTRACT: Large Language Models (LLMs) excel at intuitive, implicit reasoning. Guiding
LLMs to construct thought chains can enhance their deliberate reasoning
abilities, but also faces challenges such as hallucination. Knowledge Graphs
(KGs) can provide explicit structured knowledge for LLMs to alleviate these
issues. However, existing KG-enhanced methods often overlook explicit graph
learning, making it challenging to efficiently provide precise reasoning chains
for LLMs. Following dual-process theory, we propose Dual-Reasoning (DualR), a
novel framework that integrates an external system based on Graph Neural
Network (GNN) for explicit reasoning on KGs, complementing the implicit
reasoning of LLMs through externalized reasoning chains. DualR designs an
LLM-empowered GNN module for explicit learning on KGs, efficiently extracting
high-quality reasoning chains. These reasoning chains are then refined to a
knowledge-enhanced multiple-choice prompt, guiding a frozen LLM to reason
thoughtfully for final answer determination. Extensive experiments on three
benchmark KGQA datasets demonstrate that DualR achieves state-of-the-art
performance while maintaining high efficiency and interpretability.
| no_new_dataset | 0.946547 |
2406.10292 | Chufan Gao | Chufan Gao, Jathurshan Pradeepkumar, Trisha Das, Shivashankar Thati,
Jimeng Sun | Automatically Labeling Clinical Trial Outcomes: A Large-Scale Benchmark
for Drug Development | null | null | null | null | cs.AI cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Background The cost of drug discovery and development is substantial, with
clinical trial outcomes playing a critical role in regulatory approval and
patient care. However, access to large-scale, high-quality clinical trial
outcome data remains limited, hindering advancements in predictive modeling and
evidence-based decision-making.
Methods We present the Clinical Trial Outcome (CTO) benchmark, a fully
reproducible, large-scale repository encompassing approximately 125,000 drug
and biologics trials. CTO integrates large language model (LLM) interpretations
of publications, trial phase progression tracking, sentiment analysis from news
sources, stock price movements of trial sponsors, and additional trial-related
metrics. Furthermore, we manually annotated a dataset of clinical trials
conducted between 2020 and 2024 to enhance the quality and reliability of
outcome labels.
Results The trial outcome labels in the CTO benchmark agree strongly with
expert annotations, achieving an F1 score of 94 for Phase 3 trials and 91
across all phases. Additionally, benchmarking standard machine learning models
on our manually annotated dataset revealed distribution shifts in recent
trials, underscoring the necessity of continuously updated labeling approaches.
Conclusions By analyzing CTO's performance on recent clinical trials, we
demonstrate the ongoing need for high-quality, up-to-date trial outcome labels.
We publicly release the CTO knowledge base and annotated labels at
https://chufangao.github.io/CTOD, with regular updates to support research on
clinical trial outcomes and inform data-driven improvements in drug
development.
| [
{
"version": "v1",
"created": "Thu, 13 Jun 2024 04:23:35 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Feb 2025 20:16:56 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 02:41:55 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Gao",
"Chufan",
""
],
[
"Pradeepkumar",
"Jathurshan",
""
],
[
"Das",
"Trisha",
""
],
[
"Thati",
"Shivashankar",
""
],
[
"Sun",
"Jimeng",
""
]
]
| TITLE: Automatically Labeling Clinical Trial Outcomes: A Large-Scale Benchmark
for Drug Development
ABSTRACT: Background The cost of drug discovery and development is substantial, with
clinical trial outcomes playing a critical role in regulatory approval and
patient care. However, access to large-scale, high-quality clinical trial
outcome data remains limited, hindering advancements in predictive modeling and
evidence-based decision-making.
Methods We present the Clinical Trial Outcome (CTO) benchmark, a fully
reproducible, large-scale repository encompassing approximately 125,000 drug
and biologics trials. CTO integrates large language model (LLM) interpretations
of publications, trial phase progression tracking, sentiment analysis from news
sources, stock price movements of trial sponsors, and additional trial-related
metrics. Furthermore, we manually annotated a dataset of clinical trials
conducted between 2020 and 2024 to enhance the quality and reliability of
outcome labels.
Results The trial outcome labels in the CTO benchmark agree strongly with
expert annotations, achieving an F1 score of 94 for Phase 3 trials and 91
across all phases. Additionally, benchmarking standard machine learning models
on our manually annotated dataset revealed distribution shifts in recent
trials, underscoring the necessity of continuously updated labeling approaches.
Conclusions By analyzing CTO's performance on recent clinical trials, we
demonstrate the ongoing need for high-quality, up-to-date trial outcome labels.
We publicly release the CTO knowledge base and annotated labels at
https://chufangao.github.io/CTOD, with regular updates to support research on
clinical trial outcomes and inform data-driven improvements in drug
development.
| no_new_dataset | 0.935228 |
2406.12723 | Scott Lowe | Zahra Gharaee, Scott C. Lowe, ZeMing Gong, Pablo Millan Arias,
Nicholas Pellegrino, Austin T. Wang, Joakim Bruslund Haurum, Iuliia
Zarubiieva, Lila Kari, Dirk Steinke, Graham W. Taylor, Paul Fieguth, Angel X.
Chang | BIOSCAN-5M: A Multimodal Dataset for Insect Biodiversity | null | NeurIPS 2024 | null | null | cs.LG cs.AI cs.CV q-bio.PE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | As part of an ongoing worldwide effort to comprehend and monitor insect
biodiversity, this paper presents the BIOSCAN-5M Insect dataset to the machine
learning community and establish several benchmark tasks. BIOSCAN-5M is a
comprehensive dataset containing multi-modal information for over 5 million
insect specimens, and it significantly expands existing image-based biological
datasets by including taxonomic labels, raw nucleotide barcode sequences,
assigned barcode index numbers, geographical, and size information. We propose
three benchmark experiments to demonstrate the impact of the multi-modal data
types on the classification and clustering accuracy. First, we pretrain a
masked language model on the DNA barcode sequences of the BIOSCAN-5M dataset,
and demonstrate the impact of using this large reference library on species-
and genus-level classification performance. Second, we propose a zero-shot
transfer learning task applied to images and DNA barcodes to cluster feature
embeddings obtained from self-supervised learning, to investigate whether
meaningful clusters can be derived from these representation embeddings. Third,
we benchmark multi-modality by performing contrastive learning on DNA barcodes,
image data, and taxonomic information. This yields a general shared embedding
space enabling taxonomic classification using multiple types of information and
modalities. The code repository of the BIOSCAN-5M Insect dataset is available
at https://github.com/bioscan-ml/BIOSCAN-5M.
| [
{
"version": "v1",
"created": "Tue, 18 Jun 2024 15:45:21 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Jun 2024 04:13:11 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Jun 2024 02:00:48 GMT"
},
{
"version": "v4",
"created": "Wed, 13 Nov 2024 01:45:11 GMT"
},
{
"version": "v5",
"created": "Fri, 24 Jan 2025 23:38:09 GMT"
},
{
"version": "v6",
"created": "Sat, 1 Mar 2025 00:03:47 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Gharaee",
"Zahra",
""
],
[
"Lowe",
"Scott C.",
""
],
[
"Gong",
"ZeMing",
""
],
[
"Arias",
"Pablo Millan",
""
],
[
"Pellegrino",
"Nicholas",
""
],
[
"Wang",
"Austin T.",
""
],
[
"Haurum",
"Joakim Bruslund",
""
],
[
"Zarubiieva",
"Iuliia",
""
],
[
"Kari",
"Lila",
""
],
[
"Steinke",
"Dirk",
""
],
[
"Taylor",
"Graham W.",
""
],
[
"Fieguth",
"Paul",
""
],
[
"Chang",
"Angel X.",
""
]
]
| TITLE: BIOSCAN-5M: A Multimodal Dataset for Insect Biodiversity
ABSTRACT: As part of an ongoing worldwide effort to comprehend and monitor insect
biodiversity, this paper presents the BIOSCAN-5M Insect dataset to the machine
learning community and establish several benchmark tasks. BIOSCAN-5M is a
comprehensive dataset containing multi-modal information for over 5 million
insect specimens, and it significantly expands existing image-based biological
datasets by including taxonomic labels, raw nucleotide barcode sequences,
assigned barcode index numbers, geographical, and size information. We propose
three benchmark experiments to demonstrate the impact of the multi-modal data
types on the classification and clustering accuracy. First, we pretrain a
masked language model on the DNA barcode sequences of the BIOSCAN-5M dataset,
and demonstrate the impact of using this large reference library on species-
and genus-level classification performance. Second, we propose a zero-shot
transfer learning task applied to images and DNA barcodes to cluster feature
embeddings obtained from self-supervised learning, to investigate whether
meaningful clusters can be derived from these representation embeddings. Third,
we benchmark multi-modality by performing contrastive learning on DNA barcodes,
image data, and taxonomic information. This yields a general shared embedding
space enabling taxonomic classification using multiple types of information and
modalities. The code repository of the BIOSCAN-5M Insect dataset is available
at https://github.com/bioscan-ml/BIOSCAN-5M.
| new_dataset | 0.960137 |
2406.12753 | Zhen Huang | Zhen Huang, Zengzhi Wang, Shijie Xia, Xuefeng Li, Haoyang Zou, Ruijie
Xu, Run-Ze Fan, Lyumanshan Ye, Ethan Chern, Yixin Ye, Yikai Zhang, Yuqing
Yang, Ting Wu, Binjie Wang, Shichao Sun, Yang Xiao, Yiyuan Li, Fan Zhou,
Steffi Chern, Yiwei Qin, Yan Ma, Jiadi Su, Yixiu Liu, Yuxiang Zheng, Shaoting
Zhang, Dahua Lin, Yu Qiao, Pengfei Liu | OlympicArena: Benchmarking Multi-discipline Cognitive Reasoning for
Superintelligent AI | Accepted by NeurIPS 2024 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The evolution of Artificial Intelligence (AI) has been significantly
accelerated by advancements in Large Language Models (LLMs) and Large
Multimodal Models (LMMs), gradually showcasing potential cognitive reasoning
abilities in problem-solving and scientific discovery (i.e., AI4Science) once
exclusive to human intellect. To comprehensively evaluate current models'
performance in cognitive reasoning abilities, we introduce OlympicArena, which
includes 11,163 bilingual problems across both text-only and interleaved
text-image modalities. These challenges encompass a wide range of disciplines
spanning seven fields and 62 international Olympic competitions, rigorously
examined for data leakage. We argue that the challenges in Olympic competition
problems are ideal for evaluating AI's cognitive reasoning due to their
complexity and interdisciplinary nature, which are essential for tackling
complex scientific challenges and facilitating discoveries. Beyond evaluating
performance across various disciplines using answer-only criteria, we conduct
detailed experiments and analyses from multiple perspectives. We delve into the
models' cognitive reasoning abilities, their performance across different
modalities, and their outcomes in process-level evaluations, which are vital
for tasks requiring complex reasoning with lengthy solutions. Our extensive
evaluations reveal that even advanced models like GPT-4o only achieve a 39.97%
overall accuracy, illustrating current AI limitations in complex reasoning and
multimodal integration. Through the OlympicArena, we aim to advance AI towards
superintelligence, equipping it to address more complex challenges in science
and beyond. We also provide a comprehensive set of resources to support AI
research, including a benchmark dataset, an open-source annotation platform, a
detailed evaluation tool, and a leaderboard with automatic submission features.
| [
{
"version": "v1",
"created": "Tue, 18 Jun 2024 16:20:53 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 12:55:25 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Huang",
"Zhen",
""
],
[
"Wang",
"Zengzhi",
""
],
[
"Xia",
"Shijie",
""
],
[
"Li",
"Xuefeng",
""
],
[
"Zou",
"Haoyang",
""
],
[
"Xu",
"Ruijie",
""
],
[
"Fan",
"Run-Ze",
""
],
[
"Ye",
"Lyumanshan",
""
],
[
"Chern",
"Ethan",
""
],
[
"Ye",
"Yixin",
""
],
[
"Zhang",
"Yikai",
""
],
[
"Yang",
"Yuqing",
""
],
[
"Wu",
"Ting",
""
],
[
"Wang",
"Binjie",
""
],
[
"Sun",
"Shichao",
""
],
[
"Xiao",
"Yang",
""
],
[
"Li",
"Yiyuan",
""
],
[
"Zhou",
"Fan",
""
],
[
"Chern",
"Steffi",
""
],
[
"Qin",
"Yiwei",
""
],
[
"Ma",
"Yan",
""
],
[
"Su",
"Jiadi",
""
],
[
"Liu",
"Yixiu",
""
],
[
"Zheng",
"Yuxiang",
""
],
[
"Zhang",
"Shaoting",
""
],
[
"Lin",
"Dahua",
""
],
[
"Qiao",
"Yu",
""
],
[
"Liu",
"Pengfei",
""
]
]
| TITLE: OlympicArena: Benchmarking Multi-discipline Cognitive Reasoning for
Superintelligent AI
ABSTRACT: The evolution of Artificial Intelligence (AI) has been significantly
accelerated by advancements in Large Language Models (LLMs) and Large
Multimodal Models (LMMs), gradually showcasing potential cognitive reasoning
abilities in problem-solving and scientific discovery (i.e., AI4Science) once
exclusive to human intellect. To comprehensively evaluate current models'
performance in cognitive reasoning abilities, we introduce OlympicArena, which
includes 11,163 bilingual problems across both text-only and interleaved
text-image modalities. These challenges encompass a wide range of disciplines
spanning seven fields and 62 international Olympic competitions, rigorously
examined for data leakage. We argue that the challenges in Olympic competition
problems are ideal for evaluating AI's cognitive reasoning due to their
complexity and interdisciplinary nature, which are essential for tackling
complex scientific challenges and facilitating discoveries. Beyond evaluating
performance across various disciplines using answer-only criteria, we conduct
detailed experiments and analyses from multiple perspectives. We delve into the
models' cognitive reasoning abilities, their performance across different
modalities, and their outcomes in process-level evaluations, which are vital
for tasks requiring complex reasoning with lengthy solutions. Our extensive
evaluations reveal that even advanced models like GPT-4o only achieve a 39.97%
overall accuracy, illustrating current AI limitations in complex reasoning and
multimodal integration. Through the OlympicArena, we aim to advance AI towards
superintelligence, equipping it to address more complex challenges in science
and beyond. We also provide a comprehensive set of resources to support AI
research, including a benchmark dataset, an open-source annotation platform, a
detailed evaluation tool, and a leaderboard with automatic submission features.
| no_new_dataset | 0.61708 |
2406.18380 | Roman Bresson | Roman Bresson and Giannis Nikolentzos and George Panagopoulos and
Michail Chatzianastasis and Jun Pang and Michalis Vazirgiannis | KAGNNs: Kolmogorov-Arnold Networks meet Graph Learning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In recent years, Graph Neural Networks (GNNs) have become the de facto tool
for learning node and graph representations. Most GNNs typically consist of a
sequence of neighborhood aggregation (a.k.a., message-passing) layers, within
which the representation of each node is updated based on those of its
neighbors. The most expressive message-passing GNNs can be obtained through the
use of the sum aggregator and of MLPs for feature transformation, thanks to
their universal approximation capabilities. However, the limitations of MLPs
recently motivated the introduction of another family of universal
approximators, called Kolmogorov-Arnold Networks (KANs) which rely on a
different representation theorem. In this work, we compare the performance of
KANs against that of MLPs on graph learning tasks. We implement three new
KAN-based GNN layers, inspired respectively by the GCN, GAT and GIN layers. We
evaluate two different implementations of KANs using two distinct base families
of functions, namely B-splines and radial basis functions. We perform extensive
experiments on node classification, link prediction, graph classification and
graph regression datasets. Our results indicate that KANs are on-par with or
better than MLPs on all tasks studied in this paper. We also show that the size
and training speed of RBF-based KANs is only marginally higher than for MLPs,
making them viable alternatives. Code available at
https://github.com/RomanBresson/KAGNN.
| [
{
"version": "v1",
"created": "Wed, 26 Jun 2024 14:21:21 GMT"
},
{
"version": "v2",
"created": "Mon, 1 Jul 2024 07:13:08 GMT"
},
{
"version": "v3",
"created": "Fri, 13 Dec 2024 09:34:22 GMT"
},
{
"version": "v4",
"created": "Thu, 6 Mar 2025 10:25:17 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Bresson",
"Roman",
""
],
[
"Nikolentzos",
"Giannis",
""
],
[
"Panagopoulos",
"George",
""
],
[
"Chatzianastasis",
"Michail",
""
],
[
"Pang",
"Jun",
""
],
[
"Vazirgiannis",
"Michalis",
""
]
]
| TITLE: KAGNNs: Kolmogorov-Arnold Networks meet Graph Learning
ABSTRACT: In recent years, Graph Neural Networks (GNNs) have become the de facto tool
for learning node and graph representations. Most GNNs typically consist of a
sequence of neighborhood aggregation (a.k.a., message-passing) layers, within
which the representation of each node is updated based on those of its
neighbors. The most expressive message-passing GNNs can be obtained through the
use of the sum aggregator and of MLPs for feature transformation, thanks to
their universal approximation capabilities. However, the limitations of MLPs
recently motivated the introduction of another family of universal
approximators, called Kolmogorov-Arnold Networks (KANs) which rely on a
different representation theorem. In this work, we compare the performance of
KANs against that of MLPs on graph learning tasks. We implement three new
KAN-based GNN layers, inspired respectively by the GCN, GAT and GIN layers. We
evaluate two different implementations of KANs using two distinct base families
of functions, namely B-splines and radial basis functions. We perform extensive
experiments on node classification, link prediction, graph classification and
graph regression datasets. Our results indicate that KANs are on-par with or
better than MLPs on all tasks studied in this paper. We also show that the size
and training speed of RBF-based KANs is only marginally higher than for MLPs,
making them viable alternatives. Code available at
https://github.com/RomanBresson/KAGNN.
| no_new_dataset | 0.948585 |
2407.04287 | Alex Ergasti | Alex Ergasti, Tomaso Fontanini, Claudio Ferrari, Massimo Bertozzi,
Andrea Prati | MARS: Paying more attention to visual attributes for text-based person
search | null | null | 10.1145/3721482 | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text-based person search (TBPS) is a problem that gained significant interest
within the research community. The task is that of retrieving one or more
images of a specific individual based on a textual description. The multi-modal
nature of the task requires learning representations that bridge text and image
data within a shared latent space. Existing TBPS systems face two major
challenges. One is defined as inter-identity noise that is due to the inherent
vagueness and imprecision of text descriptions and it indicates how
descriptions of visual attributes can be generally associated to different
people; the other is the intra-identity variations, which are all those
nuisances e.g. pose, illumination, that can alter the visual appearance of the
same textual attributes for a given subject. To address these issues, this
paper presents a novel TBPS architecture named MARS
(Mae-Attribute-Relation-Sensitive), which enhances current state-of-the-art
models by introducing two key components: a Visual Reconstruction Loss and an
Attribute Loss. The former employs a Masked AutoEncoder trained to reconstruct
randomly masked image patches with the aid of the textual description. In doing
so the model is encouraged to learn more expressive representations and
textual-visual relations in the latent space. The Attribute Loss, instead,
balances the contribution of different types of attributes, defined as
adjective-noun chunks of text. This loss ensures that every attribute is taken
into consideration in the person retrieval process. Extensive experiments on
three commonly used datasets, namely CUHK-PEDES, ICFG-PEDES, and RSTPReid,
report performance improvements, with significant gains in the mean Average
Precision (mAP) metric w.r.t. the current state of the art.
| [
{
"version": "v1",
"created": "Fri, 5 Jul 2024 06:44:43 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Ergasti",
"Alex",
""
],
[
"Fontanini",
"Tomaso",
""
],
[
"Ferrari",
"Claudio",
""
],
[
"Bertozzi",
"Massimo",
""
],
[
"Prati",
"Andrea",
""
]
]
| TITLE: MARS: Paying more attention to visual attributes for text-based person
search
ABSTRACT: Text-based person search (TBPS) is a problem that gained significant interest
within the research community. The task is that of retrieving one or more
images of a specific individual based on a textual description. The multi-modal
nature of the task requires learning representations that bridge text and image
data within a shared latent space. Existing TBPS systems face two major
challenges. One is defined as inter-identity noise that is due to the inherent
vagueness and imprecision of text descriptions and it indicates how
descriptions of visual attributes can be generally associated to different
people; the other is the intra-identity variations, which are all those
nuisances e.g. pose, illumination, that can alter the visual appearance of the
same textual attributes for a given subject. To address these issues, this
paper presents a novel TBPS architecture named MARS
(Mae-Attribute-Relation-Sensitive), which enhances current state-of-the-art
models by introducing two key components: a Visual Reconstruction Loss and an
Attribute Loss. The former employs a Masked AutoEncoder trained to reconstruct
randomly masked image patches with the aid of the textual description. In doing
so the model is encouraged to learn more expressive representations and
textual-visual relations in the latent space. The Attribute Loss, instead,
balances the contribution of different types of attributes, defined as
adjective-noun chunks of text. This loss ensures that every attribute is taken
into consideration in the person retrieval process. Extensive experiments on
three commonly used datasets, namely CUHK-PEDES, ICFG-PEDES, and RSTPReid,
report performance improvements, with significant gains in the mean Average
Precision (mAP) metric w.r.t. the current state of the art.
| no_new_dataset | 0.949809 |
2407.06639 | Zihao Zhou | Zihao Zhou, Antti Aitio, David Howey | Learning Li-ion battery health and degradation modes from data with
aging-aware circuit models | 11 pages, 10 figures | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Non-invasive estimation of Li-ion battery state-of-health from operational
data is valuable for battery applications, but remains challenging. Pure
model-based methods may suffer from inaccuracy and long-term instability of
parameter estimates, whereas pure data-driven methods rely heavily on training
data quality and quantity, causing lack of generality when extrapolating to
unseen cases. We apply an aging-aware equivalent circuit model for health
estimation, combining the flexibility of data-driven techniques within a
model-based approach. A simplified electrical model with voltage source and
resistor incorporates Gaussian process regression to learn capacity fade over
time and also the dependence of resistance on operating conditions and time.
The approach was validated against two datasets and shown to give accurate
performance with less than 1% relative root mean square error (RMSE) in
capacity and less than 2% mean absolute percentage error (MAPE). Critically, we
show that the open circuit voltage versus state-of-charge function must be
accurately known, and any inaccuracies or changes in this over time strongly
influence the inferred resistance. However, this feature (or bug) may also be
used to estimate in operando differential voltage curves from operational data.
| [
{
"version": "v1",
"created": "Tue, 9 Jul 2024 08:07:39 GMT"
},
{
"version": "v2",
"created": "Wed, 10 Jul 2024 01:45:08 GMT"
},
{
"version": "v3",
"created": "Sun, 14 Jul 2024 03:24:18 GMT"
},
{
"version": "v4",
"created": "Thu, 6 Mar 2025 14:54:01 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Zhou",
"Zihao",
""
],
[
"Aitio",
"Antti",
""
],
[
"Howey",
"David",
""
]
]
| TITLE: Learning Li-ion battery health and degradation modes from data with
aging-aware circuit models
ABSTRACT: Non-invasive estimation of Li-ion battery state-of-health from operational
data is valuable for battery applications, but remains challenging. Pure
model-based methods may suffer from inaccuracy and long-term instability of
parameter estimates, whereas pure data-driven methods rely heavily on training
data quality and quantity, causing lack of generality when extrapolating to
unseen cases. We apply an aging-aware equivalent circuit model for health
estimation, combining the flexibility of data-driven techniques within a
model-based approach. A simplified electrical model with voltage source and
resistor incorporates Gaussian process regression to learn capacity fade over
time and also the dependence of resistance on operating conditions and time.
The approach was validated against two datasets and shown to give accurate
performance with less than 1% relative root mean square error (RMSE) in
capacity and less than 2% mean absolute percentage error (MAPE). Critically, we
show that the open circuit voltage versus state-of-charge function must be
accurately known, and any inaccuracies or changes in this over time strongly
influence the inferred resistance. However, this feature (or bug) may also be
used to estimate in operando differential voltage curves from operational data.
| no_new_dataset | 0.947527 |
2407.07918 | Adnane Ez-Zizi | Oladipo A. Madamidola, Felix Ngobigha and Adnane Ez-zizi | Detecting new obfuscated malware variants: A lightweight and
interpretable machine learning approach | 30 pages (excluding Appendix), 5 figures and 5 tables. Now published
in Intelligent Systems with Applications
(https://doi.org/10.1016/j.iswa.2024.200472) | Intelligent Systems with Applications, 25 (2025) | 10.1016/j.iswa.2024.200472 | null | cs.CR cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning has been successfully applied in developing malware
detection systems, with a primary focus on accuracy, and increasing attention
to reducing computational overhead and improving model interpretability.
However, an important question remains underexplored: How well can machine
learning-based models detect entirely new forms of malware not present in the
training data? In this study, we present a machine learning-based system for
detecting obfuscated malware that is not only highly accurate, lightweight and
interpretable, but also capable of successfully adapting to new types of
malware attacks. Our system is capable of detecting 15 malware subtypes despite
being exclusively trained on one malware subtype, namely the Transponder from
the Spyware family. This system was built after training 15 distinct random
forest-based models, each on a different malware subtype from the
CIC-MalMem-2022 dataset. These models were evaluated against the entire range
of malware subtypes, including all unseen malware subtypes. To maintain the
system's streamlined nature, training was confined to the top five most
important features, which also enhanced interpretability. The
Transponder-focused model exhibited high accuracy, exceeding 99.8%, with an
average processing speed of 5.7 microseconds per file. We also illustrate how
the Shapley additive explanations technique can facilitate the interpretation
of the model predictions. Our research contributes to advancing malware
detection methodologies, pioneering the feasibility of detecting obfuscated
malware by exclusively training a model on a single or a few carefully selected
malware subtypes and applying it to detect unseen subtypes.
| [
{
"version": "v1",
"created": "Sun, 7 Jul 2024 12:41:40 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 12:41:21 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Madamidola",
"Oladipo A.",
""
],
[
"Ngobigha",
"Felix",
""
],
[
"Ez-zizi",
"Adnane",
""
]
]
| TITLE: Detecting new obfuscated malware variants: A lightweight and
interpretable machine learning approach
ABSTRACT: Machine learning has been successfully applied in developing malware
detection systems, with a primary focus on accuracy, and increasing attention
to reducing computational overhead and improving model interpretability.
However, an important question remains underexplored: How well can machine
learning-based models detect entirely new forms of malware not present in the
training data? In this study, we present a machine learning-based system for
detecting obfuscated malware that is not only highly accurate, lightweight and
interpretable, but also capable of successfully adapting to new types of
malware attacks. Our system is capable of detecting 15 malware subtypes despite
being exclusively trained on one malware subtype, namely the Transponder from
the Spyware family. This system was built after training 15 distinct random
forest-based models, each on a different malware subtype from the
CIC-MalMem-2022 dataset. These models were evaluated against the entire range
of malware subtypes, including all unseen malware subtypes. To maintain the
system's streamlined nature, training was confined to the top five most
important features, which also enhanced interpretability. The
Transponder-focused model exhibited high accuracy, exceeding 99.8%, with an
average processing speed of 5.7 microseconds per file. We also illustrate how
the Shapley additive explanations technique can facilitate the interpretation
of the model predictions. Our research contributes to advancing malware
detection methodologies, pioneering the feasibility of detecting obfuscated
malware by exclusively training a model on a single or a few carefully selected
malware subtypes and applying it to detect unseen subtypes.
| no_new_dataset | 0.943243 |
2407.08974 | Xue Gong | Joshua Zhi En Tan, JunJie Wee, Xue Gong, Kelin Xia | Topology-enhanced machine learning model (Top-ML) for anticancer peptide
prediction | null | null | null | null | q-bio.QM cs.LG math.GN q-bio.BM | http://creativecommons.org/licenses/by/4.0/ | Recently, therapeutic peptides have demonstrated great promise for cancer
treatment. To explore powerful anticancer peptides, artificial intelligence
(AI)-based approaches have been developed to systematically screen potential
candidates. However, the lack of efficient featurization of peptides has become
a bottleneck for these machine-learning models. In this paper, we propose a
topology-enhanced machine learning model (Top-ML) for anticancer peptides
prediction. Our Top-ML employs peptide topological features derived from its
sequence "connection" information characterized by vector and spectral
descriptors. Our Top-ML model, employing an Extra-Trees classifier, has been
validated on the AntiCP 2.0 and mACPpred 2.0 benchmark datasets, achieving
state-of-the-art performance or results comparable to existing deep learning
models, while providing greater interpretability. Our results highlight the
potential of leveraging novel topology-based featurization to accelerate the
identification of anticancer peptides.
| [
{
"version": "v1",
"created": "Fri, 12 Jul 2024 04:04:54 GMT"
},
{
"version": "v2",
"created": "Wed, 8 Jan 2025 06:42:39 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 03:33:25 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Tan",
"Joshua Zhi En",
""
],
[
"Wee",
"JunJie",
""
],
[
"Gong",
"Xue",
""
],
[
"Xia",
"Kelin",
""
]
]
| TITLE: Topology-enhanced machine learning model (Top-ML) for anticancer peptide
prediction
ABSTRACT: Recently, therapeutic peptides have demonstrated great promise for cancer
treatment. To explore powerful anticancer peptides, artificial intelligence
(AI)-based approaches have been developed to systematically screen potential
candidates. However, the lack of efficient featurization of peptides has become
a bottleneck for these machine-learning models. In this paper, we propose a
topology-enhanced machine learning model (Top-ML) for anticancer peptides
prediction. Our Top-ML employs peptide topological features derived from its
sequence "connection" information characterized by vector and spectral
descriptors. Our Top-ML model, employing an Extra-Trees classifier, has been
validated on the AntiCP 2.0 and mACPpred 2.0 benchmark datasets, achieving
state-of-the-art performance or results comparable to existing deep learning
models, while providing greater interpretability. Our results highlight the
potential of leveraging novel topology-based featurization to accelerate the
identification of anticancer peptides.
| no_new_dataset | 0.949623 |
2407.13304 | Federico Magistri | Federico Magistri, Thomas L\"abe, Elias Marks, Sumanth Nagulavancha,
Yue Pan, Claus Smitt, Lasse Klingbeil, Michael Halstead, Heiner Kuhlmann,
Chris McCool, Jens Behley, Cyrill Stachniss | A Dataset and Benchmark for Shape Completion of Fruits for Agricultural
Robotics | null | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As the world population is expected to reach 10 billion by 2050, our
agricultural production system needs to double its productivity despite a
decline of human workforce in the agricultural sector. Autonomous robotic
systems are one promising pathway to increase productivity by taking over
labor-intensive manual tasks like fruit picking. To be effective, such systems
need to monitor and interact with plants and fruits precisely, which is
challenging due to the cluttered nature of agricultural environments causing,
for example, strong occlusions. Thus, being able to estimate the complete 3D
shapes of objects in presence of occlusions is crucial for automating
operations such as fruit harvesting. In this paper, we propose the first
publicly available 3D shape completion dataset for agricultural vision systems.
We provide an RGB-D dataset for estimating the 3D shape of fruits.
Specifically, our dataset contains RGB-D frames of single sweet peppers in lab
conditions but also in a commercial greenhouse. For each fruit, we additionally
collected high-precision point clouds that we use as ground truth. For
acquiring the ground truth shape, we developed a measuring process that allows
us to record data of real sweet pepper plants, both in the lab and in the
greenhouse with high precision, and determine the shape of the sensed fruits.
We release our dataset, consisting of almost 7,000 RGB-D frames belonging to
more than 100 different fruits. We provide segmented RGB-D frames, with camera
intrinsics to easily obtain colored point clouds, together with the
corresponding high-precision, occlusion-free point clouds obtained with a
high-precision laser scanner. We additionally enable evaluation of shape
completion approaches on a hidden test set through a public challenge on a
benchmark server.
| [
{
"version": "v1",
"created": "Thu, 18 Jul 2024 09:07:23 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Sep 2024 08:16:57 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 14:06:01 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Magistri",
"Federico",
""
],
[
"Läbe",
"Thomas",
""
],
[
"Marks",
"Elias",
""
],
[
"Nagulavancha",
"Sumanth",
""
],
[
"Pan",
"Yue",
""
],
[
"Smitt",
"Claus",
""
],
[
"Klingbeil",
"Lasse",
""
],
[
"Halstead",
"Michael",
""
],
[
"Kuhlmann",
"Heiner",
""
],
[
"McCool",
"Chris",
""
],
[
"Behley",
"Jens",
""
],
[
"Stachniss",
"Cyrill",
""
]
]
| TITLE: A Dataset and Benchmark for Shape Completion of Fruits for Agricultural
Robotics
ABSTRACT: As the world population is expected to reach 10 billion by 2050, our
agricultural production system needs to double its productivity despite a
decline of human workforce in the agricultural sector. Autonomous robotic
systems are one promising pathway to increase productivity by taking over
labor-intensive manual tasks like fruit picking. To be effective, such systems
need to monitor and interact with plants and fruits precisely, which is
challenging due to the cluttered nature of agricultural environments causing,
for example, strong occlusions. Thus, being able to estimate the complete 3D
shapes of objects in presence of occlusions is crucial for automating
operations such as fruit harvesting. In this paper, we propose the first
publicly available 3D shape completion dataset for agricultural vision systems.
We provide an RGB-D dataset for estimating the 3D shape of fruits.
Specifically, our dataset contains RGB-D frames of single sweet peppers in lab
conditions but also in a commercial greenhouse. For each fruit, we additionally
collected high-precision point clouds that we use as ground truth. For
acquiring the ground truth shape, we developed a measuring process that allows
us to record data of real sweet pepper plants, both in the lab and in the
greenhouse with high precision, and determine the shape of the sensed fruits.
We release our dataset, consisting of almost 7,000 RGB-D frames belonging to
more than 100 different fruits. We provide segmented RGB-D frames, with camera
intrinsics to easily obtain colored point clouds, together with the
corresponding high-precision, occlusion-free point clouds obtained with a
high-precision laser scanner. We additionally enable evaluation of shape
completion approaches on a hidden test set through a public challenge on a
benchmark server.
| new_dataset | 0.96395 |
2407.18125 | Vito Paolo Pastore | Roberto Di Via, Francesca Odone, Vito Paolo Pastore | Self-supervised pre-training with diffusion model for few-shot landmark
detection in x-ray images | Accepted at WACV 2025 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Deep neural networks have been extensively applied in the medical domain for
various tasks, including image classification, segmentation, and landmark
detection. However, their application is often hindered by data scarcity, both
in terms of available annotations and images. This study introduces a novel
application of denoising diffusion probabilistic models (DDPMs) to the landmark
detection task, specifically addressing the challenge of limited annotated data
in x-ray imaging. Our key innovation lies in leveraging DDPMs for
self-supervised pre-training in landmark detection, a previously unexplored
approach in this domain. This method enables accurate landmark detection with
minimal annotated training data (as few as 50 images), surpassing both ImageNet
supervised pre-training and traditional self-supervised techniques across three
popular x-ray benchmark datasets. To our knowledge, this work represents the
first application of diffusion models for self-supervised learning in landmark
detection, which may offer a valuable pre-training approach in few-shot
regimes, for mitigating data scarcity.
| [
{
"version": "v1",
"created": "Thu, 25 Jul 2024 15:32:59 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Oct 2024 16:10:10 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 17:03:35 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Di Via",
"Roberto",
""
],
[
"Odone",
"Francesca",
""
],
[
"Pastore",
"Vito Paolo",
""
]
]
| TITLE: Self-supervised pre-training with diffusion model for few-shot landmark
detection in x-ray images
ABSTRACT: Deep neural networks have been extensively applied in the medical domain for
various tasks, including image classification, segmentation, and landmark
detection. However, their application is often hindered by data scarcity, both
in terms of available annotations and images. This study introduces a novel
application of denoising diffusion probabilistic models (DDPMs) to the landmark
detection task, specifically addressing the challenge of limited annotated data
in x-ray imaging. Our key innovation lies in leveraging DDPMs for
self-supervised pre-training in landmark detection, a previously unexplored
approach in this domain. This method enables accurate landmark detection with
minimal annotated training data (as few as 50 images), surpassing both ImageNet
supervised pre-training and traditional self-supervised techniques across three
popular x-ray benchmark datasets. To our knowledge, this work represents the
first application of diffusion models for self-supervised learning in landmark
detection, which may offer a valuable pre-training approach in few-shot
regimes, for mitigating data scarcity.
| no_new_dataset | 0.953362 |
2407.18691 | Mengjie Zhao | Mengjie Zhao, Cees Taal, Stephan Baggerohr, Olga Fink | Graph Neural Networks for Virtual Sensing in Complex Systems: Addressing
Heterogeneous Temporal Dynamics | This paper extends our previous conference paper (Best Paper at
European Conference of the PHM Society 2024,
https://doi.org/10.36001/phme.2024.v8i1.3998). Accepted by Mechanical Systems
and Signal Processing (MSSP) | null | null | null | cs.LG cs.AI cs.CE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Real-time condition monitoring is crucial for the reliable and efficient
operation of complex systems. However, relying solely on physical sensors can
be limited due to their cost, placement constraints, or inability to directly
measure certain critical parameters. Virtual sensing addresses these
limitations by leveraging readily available sensor data and system knowledge to
estimate inaccessible parameters or infer system states. The increasing
complexity of industrial systems necessitates deployments of sensors with
diverse modalities to provide a comprehensive understanding of system states.
These sensors capture data at varying frequencies to monitor both rapid and
slowly varying system dynamics, as well as local and global state evolutions of
the systems. This leads to heterogeneous temporal dynamics, which, particularly
under varying operational end environmental conditions, pose a significant
challenge for accurate virtual sensing. To address this, we propose a
Heterogeneous Temporal Graph Neural Network (HTGNN) framework. HTGNN explicitly
models signals from diverse sensors and integrates operating conditions into
the model architecture. We evaluate HTGNN using two newly released datasets: a
bearing dataset with diverse load conditions for bearing load prediction and a
year-long simulated dataset for predicting bridge live loads. Our results
demonstrate that HTGNN significantly outperforms established baseline methods
in both tasks, particularly under highly varying operating conditions. These
results highlight HTGNN's potential as a robust and accurate virtual sensing
approach for complex systems, paving the way for improved monitoring,
predictive maintenance, and enhanced system performance. Our code and data are
available under https://github.com/EPFL-IMOS/htgnn.
| [
{
"version": "v1",
"created": "Fri, 26 Jul 2024 12:16:53 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 15:47:01 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Zhao",
"Mengjie",
""
],
[
"Taal",
"Cees",
""
],
[
"Baggerohr",
"Stephan",
""
],
[
"Fink",
"Olga",
""
]
]
| TITLE: Graph Neural Networks for Virtual Sensing in Complex Systems: Addressing
Heterogeneous Temporal Dynamics
ABSTRACT: Real-time condition monitoring is crucial for the reliable and efficient
operation of complex systems. However, relying solely on physical sensors can
be limited due to their cost, placement constraints, or inability to directly
measure certain critical parameters. Virtual sensing addresses these
limitations by leveraging readily available sensor data and system knowledge to
estimate inaccessible parameters or infer system states. The increasing
complexity of industrial systems necessitates deployments of sensors with
diverse modalities to provide a comprehensive understanding of system states.
These sensors capture data at varying frequencies to monitor both rapid and
slowly varying system dynamics, as well as local and global state evolutions of
the systems. This leads to heterogeneous temporal dynamics, which, particularly
under varying operational end environmental conditions, pose a significant
challenge for accurate virtual sensing. To address this, we propose a
Heterogeneous Temporal Graph Neural Network (HTGNN) framework. HTGNN explicitly
models signals from diverse sensors and integrates operating conditions into
the model architecture. We evaluate HTGNN using two newly released datasets: a
bearing dataset with diverse load conditions for bearing load prediction and a
year-long simulated dataset for predicting bridge live loads. Our results
demonstrate that HTGNN significantly outperforms established baseline methods
in both tasks, particularly under highly varying operating conditions. These
results highlight HTGNN's potential as a robust and accurate virtual sensing
approach for complex systems, paving the way for improved monitoring,
predictive maintenance, and enhanced system performance. Our code and data are
available under https://github.com/EPFL-IMOS/htgnn.
| new_dataset | 0.964921 |
2408.06707 | JunYong Choi | JunYong Choi, SeokYeong Lee, Haesol Park, Seung-Won Jung, Ig-Jae Kim,
Junghyun Cho | MAIR++: Improving Multi-view Attention Inverse Rendering with Implicit
Lighting Representation | Published in TPAMI. project page :
https://bring728.github.io/mairplusplus.project/. arXiv admin note: text
overlap with arXiv:2303.12368 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we propose a scene-level inverse rendering framework that uses
multi-view images to decompose the scene into geometry, SVBRDF, and 3D
spatially-varying lighting. While multi-view images have been widely used for
object-level inverse rendering, scene-level inverse rendering has primarily
been studied using single-view images due to the lack of a dataset containing
high dynamic range multi-view images with ground-truth geometry, material, and
spatially-varying lighting. To improve the quality of scene-level inverse
rendering, a novel framework called Multi-view Attention Inverse Rendering
(MAIR) was recently introduced. MAIR performs scene-level multi-view inverse
rendering by expanding the OpenRooms dataset, designing efficient pipelines to
handle multi-view images, and splitting spatially-varying lighting. Although
MAIR showed impressive results, its lighting representation is fixed to
spherical Gaussians, which limits its ability to render images realistically.
Consequently, MAIR cannot be directly used in applications such as material
editing. Moreover, its multi-view aggregation networks have difficulties
extracting rich features because they only focus on the mean and variance
between multi-view features. In this paper, we propose its extended version,
called MAIR++. MAIR++ addresses the aforementioned limitations by introducing
an implicit lighting representation that accurately captures the lighting
conditions of an image while facilitating realistic rendering. Furthermore, we
design a directional attention-based multi-view aggregation network to infer
more intricate relationships between views. Experimental results show that
MAIR++ not only achieves better performance than MAIR and single-view-based
methods, but also displays robust performance on unseen real-world scenes.
| [
{
"version": "v1",
"created": "Tue, 13 Aug 2024 08:04:23 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 02:03:44 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Choi",
"JunYong",
""
],
[
"Lee",
"SeokYeong",
""
],
[
"Park",
"Haesol",
""
],
[
"Jung",
"Seung-Won",
""
],
[
"Kim",
"Ig-Jae",
""
],
[
"Cho",
"Junghyun",
""
]
]
| TITLE: MAIR++: Improving Multi-view Attention Inverse Rendering with Implicit
Lighting Representation
ABSTRACT: In this paper, we propose a scene-level inverse rendering framework that uses
multi-view images to decompose the scene into geometry, SVBRDF, and 3D
spatially-varying lighting. While multi-view images have been widely used for
object-level inverse rendering, scene-level inverse rendering has primarily
been studied using single-view images due to the lack of a dataset containing
high dynamic range multi-view images with ground-truth geometry, material, and
spatially-varying lighting. To improve the quality of scene-level inverse
rendering, a novel framework called Multi-view Attention Inverse Rendering
(MAIR) was recently introduced. MAIR performs scene-level multi-view inverse
rendering by expanding the OpenRooms dataset, designing efficient pipelines to
handle multi-view images, and splitting spatially-varying lighting. Although
MAIR showed impressive results, its lighting representation is fixed to
spherical Gaussians, which limits its ability to render images realistically.
Consequently, MAIR cannot be directly used in applications such as material
editing. Moreover, its multi-view aggregation networks have difficulties
extracting rich features because they only focus on the mean and variance
between multi-view features. In this paper, we propose its extended version,
called MAIR++. MAIR++ addresses the aforementioned limitations by introducing
an implicit lighting representation that accurately captures the lighting
conditions of an image while facilitating realistic rendering. Furthermore, we
design a directional attention-based multi-view aggregation network to infer
more intricate relationships between views. Experimental results show that
MAIR++ not only achieves better performance than MAIR and single-view-based
methods, but also displays robust performance on unseen real-world scenes.
| no_new_dataset | 0.951639 |
2408.09110 | Jiancheng Pan | Jiancheng Pan, Yanxing Liu, Yuqian Fu, Muyuan Ma, Jiahao Li, Danda
Pani Paudel, Luc Van Gool and Xiaomeng Huang | Locate Anything on Earth: Advancing Open-Vocabulary Object Detection for
Remote Sensing Community | 15 pages, 11 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object detection, particularly open-vocabulary object detection, plays a
crucial role in Earth sciences, such as environmental monitoring, natural
disaster assessment, and land-use planning. However, existing open-vocabulary
detectors, primarily trained on natural-world images, struggle to generalize to
remote sensing images due to a significant data domain gap. Thus, this paper
aims to advance the development of open-vocabulary object detection in remote
sensing community. To achieve this, we first reformulate the task as Locate
Anything on Earth (LAE) with the goal of detecting any novel concepts on Earth.
We then developed the LAE-Label Engine which collects, auto-annotates, and
unifies up to 10 remote sensing datasets creating the LAE-1M - the first
large-scale remote sensing object detection dataset with broad category
coverage. Using the LAE-1M, we further propose and train the novel LAE-DINO
Model, the first open-vocabulary foundation object detector for the LAE task,
featuring Dynamic Vocabulary Construction (DVC) and Visual-Guided Text Prompt
Learning (VisGT) modules. DVC dynamically constructs vocabulary for each
training batch, while VisGT maps visual features to semantic space, enhancing
text features. We comprehensively conduct experiments on established remote
sensing benchmark DIOR, DOTAv2.0, as well as our newly introduced 80-class
LAE-80C benchmark. Results demonstrate the advantages of the LAE-1M dataset and
the effectiveness of the LAE-DINO method.
| [
{
"version": "v1",
"created": "Sat, 17 Aug 2024 06:24:43 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Feb 2025 18:01:16 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 09:26:00 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Pan",
"Jiancheng",
""
],
[
"Liu",
"Yanxing",
""
],
[
"Fu",
"Yuqian",
""
],
[
"Ma",
"Muyuan",
""
],
[
"Li",
"Jiahao",
""
],
[
"Paudel",
"Danda Pani",
""
],
[
"Van Gool",
"Luc",
""
],
[
"Huang",
"Xiaomeng",
""
]
]
| TITLE: Locate Anything on Earth: Advancing Open-Vocabulary Object Detection for
Remote Sensing Community
ABSTRACT: Object detection, particularly open-vocabulary object detection, plays a
crucial role in Earth sciences, such as environmental monitoring, natural
disaster assessment, and land-use planning. However, existing open-vocabulary
detectors, primarily trained on natural-world images, struggle to generalize to
remote sensing images due to a significant data domain gap. Thus, this paper
aims to advance the development of open-vocabulary object detection in remote
sensing community. To achieve this, we first reformulate the task as Locate
Anything on Earth (LAE) with the goal of detecting any novel concepts on Earth.
We then developed the LAE-Label Engine which collects, auto-annotates, and
unifies up to 10 remote sensing datasets creating the LAE-1M - the first
large-scale remote sensing object detection dataset with broad category
coverage. Using the LAE-1M, we further propose and train the novel LAE-DINO
Model, the first open-vocabulary foundation object detector for the LAE task,
featuring Dynamic Vocabulary Construction (DVC) and Visual-Guided Text Prompt
Learning (VisGT) modules. DVC dynamically constructs vocabulary for each
training batch, while VisGT maps visual features to semantic space, enhancing
text features. We comprehensively conduct experiments on established remote
sensing benchmark DIOR, DOTAv2.0, as well as our newly introduced 80-class
LAE-80C benchmark. Results demonstrate the advantages of the LAE-1M dataset and
the effectiveness of the LAE-DINO method.
| new_dataset | 0.859133 |
2408.14507 | Longyu Feng | Longyu Feng and Huahang Li and Chen Jason Zhang | Prompt-Matcher: Leveraging Large Models to Reduce Uncertainty in Schema
Matching Results | null | null | null | null | cs.DB cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Schema matching is the process of identifying correspondences between the
elements of two given schemata, essential for database management systems, data
integration, and data warehousing. For datasets across different scenarios, the
optimal schema matching algorithm is different. For single algorithm,
hyperparameter tuning also cases multiple results. All results assigned equal
probabilities are stored in probabilistic databases to facilitate uncertainty
management. The substantial degree of uncertainty diminishes the efficiency and
reliability of data processing, thereby precluding the provision of more
accurate information for decision-makers. To address this problem, we introduce
a new approach based on fine-grained correspondence verification with specific
prompt of Large Language Model.
Our approach is an iterative loop that consists of three main components: (1)
the correspondence selection algorithm, (2) correspondence verification, and
(3) the update of probability distribution. The core idea is that
correspondences intersect across multiple results, thereby linking the
verification of correspondences to the reduction of uncertainty in candidate
results.
The task of selecting an optimal correspondence set to maximize the
anticipated uncertainty reduction within a fixed budgetary framework is
established as an NP-hard problem. We propose a novel $(1-1/e)$-approximation
algorithm that significantly outperforms brute algorithm in terms of
computational efficiency. To enhance correspondence verification, we have
developed two prompt templates that enable GPT-4 to achieve state-of-the-art
performance across two established benchmark datasets. Our comprehensive
experimental evaluation demonstrates the superior effectiveness and robustness
of the proposed approach.
| [
{
"version": "v1",
"created": "Sat, 24 Aug 2024 16:54:08 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 11:15:58 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 10:26:32 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Feng",
"Longyu",
""
],
[
"Li",
"Huahang",
""
],
[
"Zhang",
"Chen Jason",
""
]
]
| TITLE: Prompt-Matcher: Leveraging Large Models to Reduce Uncertainty in Schema
Matching Results
ABSTRACT: Schema matching is the process of identifying correspondences between the
elements of two given schemata, essential for database management systems, data
integration, and data warehousing. For datasets across different scenarios, the
optimal schema matching algorithm is different. For single algorithm,
hyperparameter tuning also cases multiple results. All results assigned equal
probabilities are stored in probabilistic databases to facilitate uncertainty
management. The substantial degree of uncertainty diminishes the efficiency and
reliability of data processing, thereby precluding the provision of more
accurate information for decision-makers. To address this problem, we introduce
a new approach based on fine-grained correspondence verification with specific
prompt of Large Language Model.
Our approach is an iterative loop that consists of three main components: (1)
the correspondence selection algorithm, (2) correspondence verification, and
(3) the update of probability distribution. The core idea is that
correspondences intersect across multiple results, thereby linking the
verification of correspondences to the reduction of uncertainty in candidate
results.
The task of selecting an optimal correspondence set to maximize the
anticipated uncertainty reduction within a fixed budgetary framework is
established as an NP-hard problem. We propose a novel $(1-1/e)$-approximation
algorithm that significantly outperforms brute algorithm in terms of
computational efficiency. To enhance correspondence verification, we have
developed two prompt templates that enable GPT-4 to achieve state-of-the-art
performance across two established benchmark datasets. Our comprehensive
experimental evaluation demonstrates the superior effectiveness and robustness
of the proposed approach.
| no_new_dataset | 0.943815 |
2409.01329 | Lucas Lange | Lucas Lange and Maurice-Maximilian Heykeroth and Erhard Rahm | Assessing the Impact of Image Dataset Features on Privacy-Preserving
Machine Learning | Accepted at 21st Conference on Database Systems for Business,
Technology and Web (BTW 2025) | 21st Conference on Database Systems for Business, Technology and
Web (BTW 2025) | 10.18420/BTW2025-27 | null | cs.LG cs.CR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine Learning (ML) is crucial in many sectors, including computer vision.
However, ML models trained on sensitive data face security challenges, as they
can be attacked and leak information. Privacy-Preserving Machine Learning
(PPML) addresses this by using Differential Privacy (DP) to balance utility and
privacy. This study identifies image dataset characteristics that affect the
utility and vulnerability of private and non-private Convolutional Neural
Network (CNN) models. Through analyzing multiple datasets and privacy budgets,
we find that imbalanced datasets increase vulnerability in minority classes,
but DP mitigates this issue. Datasets with fewer classes improve both model
utility and privacy, while high entropy or low Fisher Discriminant Ratio (FDR)
datasets deteriorate the utility-privacy trade-off. These insights offer
valuable guidance for practitioners and researchers in estimating and
optimizing the utility-privacy trade-off in image datasets, helping to inform
data and privacy modifications for better outcomes based on dataset
characteristics.
| [
{
"version": "v1",
"created": "Mon, 2 Sep 2024 15:30:27 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Dec 2024 14:15:21 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Lange",
"Lucas",
""
],
[
"Heykeroth",
"Maurice-Maximilian",
""
],
[
"Rahm",
"Erhard",
""
]
]
| TITLE: Assessing the Impact of Image Dataset Features on Privacy-Preserving
Machine Learning
ABSTRACT: Machine Learning (ML) is crucial in many sectors, including computer vision.
However, ML models trained on sensitive data face security challenges, as they
can be attacked and leak information. Privacy-Preserving Machine Learning
(PPML) addresses this by using Differential Privacy (DP) to balance utility and
privacy. This study identifies image dataset characteristics that affect the
utility and vulnerability of private and non-private Convolutional Neural
Network (CNN) models. Through analyzing multiple datasets and privacy budgets,
we find that imbalanced datasets increase vulnerability in minority classes,
but DP mitigates this issue. Datasets with fewer classes improve both model
utility and privacy, while high entropy or low Fisher Discriminant Ratio (FDR)
datasets deteriorate the utility-privacy trade-off. These insights offer
valuable guidance for practitioners and researchers in estimating and
optimizing the utility-privacy trade-off in image datasets, helping to inform
data and privacy modifications for better outcomes based on dataset
characteristics.
| no_new_dataset | 0.951863 |
2409.02421 | Jiatao Chen | Jiatao Chen, Tianming Xie, Xing Tang, Jing Wang, Wenjing Dong, Bing
Shi | MusicMamba: A Dual-Feature Modeling Approach for Generating Chinese
Traditional Music with Modal Precision | Accepted by ICASSP 2025 | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, deep learning has significantly advanced the MIDI domain,
solidifying music generation as a key application of artificial intelligence.
However, existing research primarily focuses on Western music and encounters
challenges in generating melodies for Chinese traditional music, especially in
capturing modal characteristics and emotional expression. To address these
issues, we propose a new architecture, the Dual-Feature Modeling Module, which
integrates the long-range dependency modeling of the Mamba Block with the
global structure capturing capabilities of the Transformer Block. Additionally,
we introduce the Bidirectional Mamba Fusion Layer, which integrates local
details and global structures through bidirectional scanning, enhancing the
modeling of complex sequences. Building on this architecture, we propose the
REMI-M representation, which more accurately captures and generates modal
information in melodies. To support this research, we developed FolkDB, a
high-quality Chinese traditional music dataset encompassing various styles and
totaling over 11 hours of music. Experimental results demonstrate that the
proposed architecture excels in generating melodies with Chinese traditional
music characteristics, offering a new and effective solution for music
generation.
| [
{
"version": "v1",
"created": "Wed, 4 Sep 2024 04:00:22 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 02:11:09 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Chen",
"Jiatao",
""
],
[
"Xie",
"Tianming",
""
],
[
"Tang",
"Xing",
""
],
[
"Wang",
"Jing",
""
],
[
"Dong",
"Wenjing",
""
],
[
"Shi",
"Bing",
""
]
]
| TITLE: MusicMamba: A Dual-Feature Modeling Approach for Generating Chinese
Traditional Music with Modal Precision
ABSTRACT: In recent years, deep learning has significantly advanced the MIDI domain,
solidifying music generation as a key application of artificial intelligence.
However, existing research primarily focuses on Western music and encounters
challenges in generating melodies for Chinese traditional music, especially in
capturing modal characteristics and emotional expression. To address these
issues, we propose a new architecture, the Dual-Feature Modeling Module, which
integrates the long-range dependency modeling of the Mamba Block with the
global structure capturing capabilities of the Transformer Block. Additionally,
we introduce the Bidirectional Mamba Fusion Layer, which integrates local
details and global structures through bidirectional scanning, enhancing the
modeling of complex sequences. Building on this architecture, we propose the
REMI-M representation, which more accurately captures and generates modal
information in melodies. To support this research, we developed FolkDB, a
high-quality Chinese traditional music dataset encompassing various styles and
totaling over 11 hours of music. Experimental results demonstrate that the
proposed architecture excels in generating melodies with Chinese traditional
music characteristics, offering a new and effective solution for music
generation.
| new_dataset | 0.957078 |
2409.07055 | Shuyuan Zheng | Junkai Liu, Yujie Tong, Hui Huang, Bowen Zheng, Yiran Hu, Peicheng Wu,
Chuan Xiao, Makoto Onizuka, Muyun Yang, Shuyuan Zheng | Legal Fact Prediction: The Missing Piece in Legal Judgment Prediction | null | null | null | null | cs.CL cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Legal judgment prediction (LJP), which enables litigants and their lawyers to
forecast judgment outcomes and refine litigation strategies, has emerged as a
crucial legal NLP task. Existing studies typically utilize legal facts, i.e.,
facts that have been established by evidence and determined by the judge, to
predict the judgment. However, legal facts are often difficult to obtain in the
early stages of litigation, significantly limiting the practical applicability
of fact-based LJP. To address this limitation, we propose a novel legal NLP
task: \textit{legal fact prediction} (LFP), which takes the evidence submitted
by litigants for trial as input to predict legal facts, thereby empowering
fact-based LJP technologies to perform prediction in the absence of
ground-truth legal facts. We also propose the first benchmark dataset,
LFPBench, for evaluating the LFP task. Our extensive experiments on LFPBench
demonstrate the effectiveness of LFP-empowered LJP and highlight promising
research directions for LFP. Our code and data are available at
https://github.com/HPRCEST/LFPBench.
| [
{
"version": "v1",
"created": "Wed, 11 Sep 2024 07:01:08 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 05:48:54 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Liu",
"Junkai",
""
],
[
"Tong",
"Yujie",
""
],
[
"Huang",
"Hui",
""
],
[
"Zheng",
"Bowen",
""
],
[
"Hu",
"Yiran",
""
],
[
"Wu",
"Peicheng",
""
],
[
"Xiao",
"Chuan",
""
],
[
"Onizuka",
"Makoto",
""
],
[
"Yang",
"Muyun",
""
],
[
"Zheng",
"Shuyuan",
""
]
]
| TITLE: Legal Fact Prediction: The Missing Piece in Legal Judgment Prediction
ABSTRACT: Legal judgment prediction (LJP), which enables litigants and their lawyers to
forecast judgment outcomes and refine litigation strategies, has emerged as a
crucial legal NLP task. Existing studies typically utilize legal facts, i.e.,
facts that have been established by evidence and determined by the judge, to
predict the judgment. However, legal facts are often difficult to obtain in the
early stages of litigation, significantly limiting the practical applicability
of fact-based LJP. To address this limitation, we propose a novel legal NLP
task: \textit{legal fact prediction} (LFP), which takes the evidence submitted
by litigants for trial as input to predict legal facts, thereby empowering
fact-based LJP technologies to perform prediction in the absence of
ground-truth legal facts. We also propose the first benchmark dataset,
LFPBench, for evaluating the LFP task. Our extensive experiments on LFPBench
demonstrate the effectiveness of LFP-empowered LJP and highlight promising
research directions for LFP. Our code and data are available at
https://github.com/HPRCEST/LFPBench.
| new_dataset | 0.96682 |
2409.07723 | Bojian Li | Bojian Li, Bo Liu, Xinning Yao, Jinghua Yue, Fugen Zhou | Advancing Depth Anything Model for Unsupervised Monocular Depth
Estimation in Endoscopy | 8 pages, 7 figures | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Depth estimation is a cornerstone of 3D reconstruction and plays a vital role
in minimally invasive endoscopic surgeries. However, most current depth
estimation networks rely on traditional convolutional neural networks, which
are limited in their ability to capture global information. Foundation models
offer a promising approach to enhance depth estimation, but those models
currently available are primarily trained on natural images, leading to
suboptimal performance when applied to endoscopic images. In this work, we
introduce a novel fine-tuning strategy for the Depth Anything Model and
integrate it with an intrinsic-based unsupervised monocular depth estimation
framework. Our approach includes a low-rank adaptation technique based on
random vectors, which improves the model's adaptability to different scales.
Additionally, we propose a residual block built on depthwise separable
convolution to compensate for the transformer's limited ability to capture
local features. Our experimental results on the SCARED dataset and Hamlyn
dataset show that our method achieves state-of-the-art performance while
minimizing the number of trainable parameters. Applying this method in
minimally invasive endoscopic surgery can enhance surgeons' spatial awareness,
thereby improving the precision and safety of the procedures.
| [
{
"version": "v1",
"created": "Thu, 12 Sep 2024 03:04:43 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 01:40:10 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Li",
"Bojian",
""
],
[
"Liu",
"Bo",
""
],
[
"Yao",
"Xinning",
""
],
[
"Yue",
"Jinghua",
""
],
[
"Zhou",
"Fugen",
""
]
]
| TITLE: Advancing Depth Anything Model for Unsupervised Monocular Depth
Estimation in Endoscopy
ABSTRACT: Depth estimation is a cornerstone of 3D reconstruction and plays a vital role
in minimally invasive endoscopic surgeries. However, most current depth
estimation networks rely on traditional convolutional neural networks, which
are limited in their ability to capture global information. Foundation models
offer a promising approach to enhance depth estimation, but those models
currently available are primarily trained on natural images, leading to
suboptimal performance when applied to endoscopic images. In this work, we
introduce a novel fine-tuning strategy for the Depth Anything Model and
integrate it with an intrinsic-based unsupervised monocular depth estimation
framework. Our approach includes a low-rank adaptation technique based on
random vectors, which improves the model's adaptability to different scales.
Additionally, we propose a residual block built on depthwise separable
convolution to compensate for the transformer's limited ability to capture
local features. Our experimental results on the SCARED dataset and Hamlyn
dataset show that our method achieves state-of-the-art performance while
minimizing the number of trainable parameters. Applying this method in
minimally invasive endoscopic surgery can enhance surgeons' spatial awareness,
thereby improving the precision and safety of the procedures.
| no_new_dataset | 0.94625 |
2409.08388 | Hossein Resani | Hossein Resani, Behrooz Nasihatkon, Mohammadreza Alimoradi Jazi | Continual Learning in 3D Point Clouds: Employing Spectral Techniques for
Exemplar Selection | Accepted to WACV 2025, Tucson, Arizona, USA | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce a novel framework for Continual Learning in 3D object
classification. Our approach, CL3D, is based on the selection of prototypes
from each class using spectral clustering. For non-Euclidean data such as point
clouds, spectral clustering can be employed as long as one can define a
distance measure between pairs of samples. Choosing the appropriate distance
measure enables us to leverage 3D geometric characteristics to identify
representative prototypes for each class. We explore the effectiveness of
clustering in the input space (3D points), local feature space
(1024-dimensional points), and global feature space. We conduct experiments on
the ModelNet40, ShapeNet, and ScanNet datasets, achieving state-of-the-art
accuracy exclusively through the use of input space features. By leveraging the
combined input, local, and global features, we have improved the
state-of-the-art on ModelNet and ShapeNet, utilizing nearly half the memory
used by competing approaches. For the challenging ScanNet dataset, our method
enhances accuracy by 4.1% while consuming just 28% of the memory used by our
competitors, demonstrating the scalability of our approach.
| [
{
"version": "v1",
"created": "Thu, 12 Sep 2024 20:34:34 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 13:19:58 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Resani",
"Hossein",
""
],
[
"Nasihatkon",
"Behrooz",
""
],
[
"Jazi",
"Mohammadreza Alimoradi",
""
]
]
| TITLE: Continual Learning in 3D Point Clouds: Employing Spectral Techniques for
Exemplar Selection
ABSTRACT: We introduce a novel framework for Continual Learning in 3D object
classification. Our approach, CL3D, is based on the selection of prototypes
from each class using spectral clustering. For non-Euclidean data such as point
clouds, spectral clustering can be employed as long as one can define a
distance measure between pairs of samples. Choosing the appropriate distance
measure enables us to leverage 3D geometric characteristics to identify
representative prototypes for each class. We explore the effectiveness of
clustering in the input space (3D points), local feature space
(1024-dimensional points), and global feature space. We conduct experiments on
the ModelNet40, ShapeNet, and ScanNet datasets, achieving state-of-the-art
accuracy exclusively through the use of input space features. By leveraging the
combined input, local, and global features, we have improved the
state-of-the-art on ModelNet and ShapeNet, utilizing nearly half the memory
used by competing approaches. For the challenging ScanNet dataset, our method
enhances accuracy by 4.1% while consuming just 28% of the memory used by our
competitors, demonstrating the scalability of our approach.
| no_new_dataset | 0.949576 |
2409.08824 | Kaijie Yin | Kaijie Yin and Tian Gao and Hui Kong | Pathfinder for Low-altitude Aircraft with Binary Neural Network | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A prior global topological map (e.g., the OpenStreetMap, OSM) can boost the
performance of autonomous mapping by a ground mobile robot. However, the prior
map is usually incomplete due to lacking labeling in partial paths. To solve
this problem, this paper proposes an OSM maker using airborne sensors carried
by low-altitude aircraft, where the core of the OSM maker is a novel efficient
pathfinder approach based on LiDAR and camera data, i.e., a binary dual-stream
road segmentation model. Specifically, a multi-scale feature extraction based
on the UNet architecture is implemented for images and point clouds. To reduce
the effect caused by the sparsity of point cloud, an attention-guided gated
block is designed to integrate image and point-cloud features. To optimize the
model for edge deployment that significantly reduces storage footprint and
computational demands, we propose a binarization streamline to each model
component, including a variant of vision transformer (ViT) architecture as the
encoder of the image branch, and new focal and perception losses to optimize
the model training. The experimental results on two datasets demonstrate that
our pathfinder method achieves SOTA accuracy with high efficiency in finding
paths from the low-level airborne sensors, and we can create complete OSM prior
maps based on the segmented road skeletons. Code and data are available at:
\href{https://github.com/IMRL/Pathfinder}{https://github.com/IMRL/Pathfinder}.
| [
{
"version": "v1",
"created": "Fri, 13 Sep 2024 13:37:33 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Sep 2024 01:29:58 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 12:26:08 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Yin",
"Kaijie",
""
],
[
"Gao",
"Tian",
""
],
[
"Kong",
"Hui",
""
]
]
| TITLE: Pathfinder for Low-altitude Aircraft with Binary Neural Network
ABSTRACT: A prior global topological map (e.g., the OpenStreetMap, OSM) can boost the
performance of autonomous mapping by a ground mobile robot. However, the prior
map is usually incomplete due to lacking labeling in partial paths. To solve
this problem, this paper proposes an OSM maker using airborne sensors carried
by low-altitude aircraft, where the core of the OSM maker is a novel efficient
pathfinder approach based on LiDAR and camera data, i.e., a binary dual-stream
road segmentation model. Specifically, a multi-scale feature extraction based
on the UNet architecture is implemented for images and point clouds. To reduce
the effect caused by the sparsity of point cloud, an attention-guided gated
block is designed to integrate image and point-cloud features. To optimize the
model for edge deployment that significantly reduces storage footprint and
computational demands, we propose a binarization streamline to each model
component, including a variant of vision transformer (ViT) architecture as the
encoder of the image branch, and new focal and perception losses to optimize
the model training. The experimental results on two datasets demonstrate that
our pathfinder method achieves SOTA accuracy with high efficiency in finding
paths from the low-level airborne sensors, and we can create complete OSM prior
maps based on the segmented road skeletons. Code and data are available at:
\href{https://github.com/IMRL/Pathfinder}{https://github.com/IMRL/Pathfinder}.
| no_new_dataset | 0.950686 |
2409.10329 | {\L}ukasz Struski | {\L}ukasz Struski, Dawid Rymarczyk, Jacek Tabor | InfoDisent: Explainability of Image Classification Models by Information
Disentanglement | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | In this work, we introduce InfoDisent, a hybrid approach to explainability
based on the information bottleneck principle. InfoDisent enables the
disentanglement of information in the final layer of any pretrained model into
atomic concepts, which can be interpreted as prototypical parts. This approach
merges the flexibility of post-hoc methods with the concept-level modeling
capabilities of self-explainable neural networks, such as ProtoPNets. We
demonstrate the effectiveness of InfoDisent through computational experiments
and user studies across various datasets using modern backbones such as ViTs
and convolutional networks. Notably, InfoDisent generalizes the prototypical
parts approach to novel domains (ImageNet).
| [
{
"version": "v1",
"created": "Mon, 16 Sep 2024 14:39:15 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 12:16:09 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Struski",
"Łukasz",
""
],
[
"Rymarczyk",
"Dawid",
""
],
[
"Tabor",
"Jacek",
""
]
]
| TITLE: InfoDisent: Explainability of Image Classification Models by Information
Disentanglement
ABSTRACT: In this work, we introduce InfoDisent, a hybrid approach to explainability
based on the information bottleneck principle. InfoDisent enables the
disentanglement of information in the final layer of any pretrained model into
atomic concepts, which can be interpreted as prototypical parts. This approach
merges the flexibility of post-hoc methods with the concept-level modeling
capabilities of self-explainable neural networks, such as ProtoPNets. We
demonstrate the effectiveness of InfoDisent through computational experiments
and user studies across various datasets using modern backbones such as ViTs
and convolutional networks. Notably, InfoDisent generalizes the prototypical
parts approach to novel domains (ImageNet).
| no_new_dataset | 0.944791 |
2409.11699 | Hubert Pham | Liam Hebert, Marialena Kyriakidi, Hubert Pham, Krishna Sayana, James
Pine, Sukhdeep Sodhi, Ambarish Jash | FLARE: Fusing Language Models and Collaborative Architectures for
Recommender Enhancement | null | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by/4.0/ | Recent proposals in recommender systems represent items with their textual
description, using a large language model. They show better results on standard
benchmarks compared to an item ID-only model, such as Bert4Rec. In this work,
we revisit the often-used Bert4Rec baseline and show that with further tuning,
Bert4Rec significantly outperforms previously reported numbers, and in some
datasets, is competitive with state-of-the-art models.
With revised baselines for item ID-only models, this paper also establishes
new competitive results for architectures that combine IDs and textual
descriptions. We demonstrate this with Flare (Fusing Language models and
collaborative Architectures for Recommender Enhancement). Flare is a novel
hybrid sequence recommender that integrates a language model with a
collaborative filtering model using a Perceiver network.
Prior studies focus evaluation on datasets with limited-corpus size, but many
commercially-applicable recommender systems common on the web must handle
larger corpora. We evaluate Flare on a more realistic dataset with a
significantly larger item vocabulary, introducing new baselines for this
setting. This paper also showcases Flare's inherent ability to support
critiquing, enabling users to provide feedback and refine recommendations. We
leverage critiquing as an evaluation method to assess the model's language
understanding and its transferability to the recommendation task.
| [
{
"version": "v1",
"created": "Wed, 18 Sep 2024 04:43:41 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 23:46:26 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Hebert",
"Liam",
""
],
[
"Kyriakidi",
"Marialena",
""
],
[
"Pham",
"Hubert",
""
],
[
"Sayana",
"Krishna",
""
],
[
"Pine",
"James",
""
],
[
"Sodhi",
"Sukhdeep",
""
],
[
"Jash",
"Ambarish",
""
]
]
| TITLE: FLARE: Fusing Language Models and Collaborative Architectures for
Recommender Enhancement
ABSTRACT: Recent proposals in recommender systems represent items with their textual
description, using a large language model. They show better results on standard
benchmarks compared to an item ID-only model, such as Bert4Rec. In this work,
we revisit the often-used Bert4Rec baseline and show that with further tuning,
Bert4Rec significantly outperforms previously reported numbers, and in some
datasets, is competitive with state-of-the-art models.
With revised baselines for item ID-only models, this paper also establishes
new competitive results for architectures that combine IDs and textual
descriptions. We demonstrate this with Flare (Fusing Language models and
collaborative Architectures for Recommender Enhancement). Flare is a novel
hybrid sequence recommender that integrates a language model with a
collaborative filtering model using a Perceiver network.
Prior studies focus evaluation on datasets with limited-corpus size, but many
commercially-applicable recommender systems common on the web must handle
larger corpora. We evaluate Flare on a more realistic dataset with a
significantly larger item vocabulary, introducing new baselines for this
setting. This paper also showcases Flare's inherent ability to support
critiquing, enabling users to provide feedback and refine recommendations. We
leverage critiquing as an evaluation method to assess the model's language
understanding and its transferability to the recommendation task.
| no_new_dataset | 0.949342 |
2409.11919 | Amaia Cardiel | Amaia Cardiel, Eloi Zablocki, Elias Ramzi, Oriane Sim\'eoni, Matthieu
Cord | LLM-wrapper: Black-Box Semantic-Aware Adaptation of Vision-Language
Models for Referring Expression Comprehension | LLM-wrapper (v3) is published as a conference paper at ICLR 2025. (v1
was presented at EVAL-FoMo workshop, ECCV 2024.) | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Vision Language Models (VLMs) have demonstrated remarkable capabilities in
various open-vocabulary tasks, yet their zero-shot performance lags behind
task-specific fine-tuned models, particularly in complex tasks like Referring
Expression Comprehension (REC). Fine-tuning usually requires 'white-box' access
to the model's architecture and weights, which is not always feasible due to
proprietary or privacy concerns. In this work, we propose LLM-wrapper, a method
for 'black-box' adaptation of VLMs for the REC task using Large Language Models
(LLMs). LLM-wrapper capitalizes on the reasoning abilities of LLMs, improved
with a light fine-tuning, to select the most relevant bounding box matching the
referring expression, from candidates generated by a zero-shot black-box VLM.
Our approach offers several advantages: it enables the adaptation of
closed-source models without needing access to their internal workings, it is
versatile as it works with any VLM, it transfers to new VLMs and datasets, and
it allows for the adaptation of an ensemble of VLMs. We evaluate LLM-wrapper on
multiple datasets using different VLMs and LLMs, demonstrating significant
performance improvements and highlighting the versatility of our method. While
LLM-wrapper is not meant to directly compete with standard white-box
fine-tuning, it offers a practical and effective alternative for black-box VLM
adaptation. Code and checkpoints are available at
https://github.com/valeoai/LLM_wrapper .
| [
{
"version": "v1",
"created": "Wed, 18 Sep 2024 12:32:25 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Oct 2024 14:52:55 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 17:12:48 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Cardiel",
"Amaia",
""
],
[
"Zablocki",
"Eloi",
""
],
[
"Ramzi",
"Elias",
""
],
[
"Siméoni",
"Oriane",
""
],
[
"Cord",
"Matthieu",
""
]
]
| TITLE: LLM-wrapper: Black-Box Semantic-Aware Adaptation of Vision-Language
Models for Referring Expression Comprehension
ABSTRACT: Vision Language Models (VLMs) have demonstrated remarkable capabilities in
various open-vocabulary tasks, yet their zero-shot performance lags behind
task-specific fine-tuned models, particularly in complex tasks like Referring
Expression Comprehension (REC). Fine-tuning usually requires 'white-box' access
to the model's architecture and weights, which is not always feasible due to
proprietary or privacy concerns. In this work, we propose LLM-wrapper, a method
for 'black-box' adaptation of VLMs for the REC task using Large Language Models
(LLMs). LLM-wrapper capitalizes on the reasoning abilities of LLMs, improved
with a light fine-tuning, to select the most relevant bounding box matching the
referring expression, from candidates generated by a zero-shot black-box VLM.
Our approach offers several advantages: it enables the adaptation of
closed-source models without needing access to their internal workings, it is
versatile as it works with any VLM, it transfers to new VLMs and datasets, and
it allows for the adaptation of an ensemble of VLMs. We evaluate LLM-wrapper on
multiple datasets using different VLMs and LLMs, demonstrating significant
performance improvements and highlighting the versatility of our method. While
LLM-wrapper is not meant to directly compete with standard white-box
fine-tuning, it offers a practical and effective alternative for black-box VLM
adaptation. Code and checkpoints are available at
https://github.com/valeoai/LLM_wrapper .
| no_new_dataset | 0.951006 |
2410.00299 | Zhangshuo Qi | Zhangshuo Qi, Junyi Ma, Jingyi Xu, Zijie Zhou, Luqi Cheng, and
Guangming Xiong | GSPR: Multimodal Place Recognition Using 3D Gaussian Splatting for
Autonomous Driving | 8 pages, 6 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Place recognition is a crucial component that enables autonomous vehicles to
obtain localization results in GPS-denied environments. In recent years,
multimodal place recognition methods have gained increasing attention. They
overcome the weaknesses of unimodal sensor systems by leveraging complementary
information from different modalities. However, most existing methods explore
cross-modality correlations through feature-level or descriptor-level fusion,
suffering from a lack of interpretability. Conversely, the recently proposed 3D
Gaussian Splatting provides a new perspective on multimodal fusion by
harmonizing different modalities into an explicit scene representation. In this
paper, we propose a 3D Gaussian Splatting-based multimodal place recognition
network dubbed GSPR. It explicitly combines multi-view RGB images and LiDAR
point clouds into a spatio-temporally unified scene representation with the
proposed Multimodal Gaussian Splatting. A network composed of 3D graph
convolution and transformer is designed to extract spatio-temporal features and
global descriptors from the Gaussian scenes for place recognition. Extensive
evaluations on three datasets demonstrate that our method can effectively
leverage complementary strengths of both multi-view cameras and LiDAR,
achieving SOTA place recognition performance while maintaining solid
generalization ability. Our open-source code will be released at
https://github.com/QiZS-BIT/GSPR.
| [
{
"version": "v1",
"created": "Tue, 1 Oct 2024 00:43:45 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 15:32:33 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Qi",
"Zhangshuo",
""
],
[
"Ma",
"Junyi",
""
],
[
"Xu",
"Jingyi",
""
],
[
"Zhou",
"Zijie",
""
],
[
"Cheng",
"Luqi",
""
],
[
"Xiong",
"Guangming",
""
]
]
| TITLE: GSPR: Multimodal Place Recognition Using 3D Gaussian Splatting for
Autonomous Driving
ABSTRACT: Place recognition is a crucial component that enables autonomous vehicles to
obtain localization results in GPS-denied environments. In recent years,
multimodal place recognition methods have gained increasing attention. They
overcome the weaknesses of unimodal sensor systems by leveraging complementary
information from different modalities. However, most existing methods explore
cross-modality correlations through feature-level or descriptor-level fusion,
suffering from a lack of interpretability. Conversely, the recently proposed 3D
Gaussian Splatting provides a new perspective on multimodal fusion by
harmonizing different modalities into an explicit scene representation. In this
paper, we propose a 3D Gaussian Splatting-based multimodal place recognition
network dubbed GSPR. It explicitly combines multi-view RGB images and LiDAR
point clouds into a spatio-temporally unified scene representation with the
proposed Multimodal Gaussian Splatting. A network composed of 3D graph
convolution and transformer is designed to extract spatio-temporal features and
global descriptors from the Gaussian scenes for place recognition. Extensive
evaluations on three datasets demonstrate that our method can effectively
leverage complementary strengths of both multi-view cameras and LiDAR,
achieving SOTA place recognition performance while maintaining solid
generalization ability. Our open-source code will be released at
https://github.com/QiZS-BIT/GSPR.
| no_new_dataset | 0.947039 |
2410.00862 | Lorenzo Cazzaro | Stefano Calzavara, Lorenzo Cazzaro, Massimo Vettori | Timber! Poisoning Decision Trees | This work has been accepted for publication in the 3rd IEEE
Conference on Secure and Trustworthy Machine Learning (IEEE SaTML 2025). The
final version will be available on IEEE Xplore. 17 pages, 7 figures, 5 tables | null | null | null | cs.LG cs.CR stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present Timber, the first white-box poisoning attack targeting decision
trees. Timber is based on a greedy attack strategy that leverages sub-tree
retraining to efficiently estimate the damage caused by poisoning a given
training instance. The attack relies on a tree annotation procedure, which
enables the sorting of training instances so that they are processed in
increasing order of the computational cost of sub-tree retraining. This sorting
yields a variant of Timber that supports an early stopping criterion, designed
to make poisoning attacks more efficient and feasible on larger datasets. We
also discuss an extension of Timber to traditional random forest models, which
is valuable since decision trees are typically combined into ensembles to
improve their predictive power. Our experimental evaluation on public datasets
demonstrates that our attacks outperform existing baselines in terms of
effectiveness, efficiency, or both. Moreover, we show that two representative
defenses can mitigate the effect of our attacks, but fail to effectively thwart
them.
| [
{
"version": "v1",
"created": "Tue, 1 Oct 2024 16:58:54 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 22:12:27 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Calzavara",
"Stefano",
""
],
[
"Cazzaro",
"Lorenzo",
""
],
[
"Vettori",
"Massimo",
""
]
]
| TITLE: Timber! Poisoning Decision Trees
ABSTRACT: We present Timber, the first white-box poisoning attack targeting decision
trees. Timber is based on a greedy attack strategy that leverages sub-tree
retraining to efficiently estimate the damage caused by poisoning a given
training instance. The attack relies on a tree annotation procedure, which
enables the sorting of training instances so that they are processed in
increasing order of the computational cost of sub-tree retraining. This sorting
yields a variant of Timber that supports an early stopping criterion, designed
to make poisoning attacks more efficient and feasible on larger datasets. We
also discuss an extension of Timber to traditional random forest models, which
is valuable since decision trees are typically combined into ensembles to
improve their predictive power. Our experimental evaluation on public datasets
demonstrates that our attacks outperform existing baselines in terms of
effectiveness, efficiency, or both. Moreover, we show that two representative
defenses can mitigate the effect of our attacks, but fail to effectively thwart
them.
| no_new_dataset | 0.945801 |
2410.01166 | Zhijian Li | Zhijian Li, Stefan Larson, Kevin Leach | Document Classification using File Names | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Rapid document classification is critical in several time-sensitive
applications like digital forensics and large-scale media classification.
Traditional approaches that rely on heavy-duty deep learning models fall short
due to high inference times over vast input datasets and computational
resources associated with analyzing whole documents. In this paper, we present
a method using lightweight supervised learning models, combined with a TF-IDF
feature extraction-based tokenization method, to accurately and efficiently
classify documents based solely on file names, that substantially reduces
inference time. Our results indicate that file name classifiers can process
more than 90% of in-scope documents with 99.63% and 96.57% accuracy when tested
on two datasets, while being 442x faster than more complex models such as DiT.
Our method offers a crucial solution to efficiently process vast document
datasets in critical scenarios, enabling fast and more reliable document
classification.
| [
{
"version": "v1",
"created": "Wed, 2 Oct 2024 01:42:19 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 19:11:50 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Li",
"Zhijian",
""
],
[
"Larson",
"Stefan",
""
],
[
"Leach",
"Kevin",
""
]
]
| TITLE: Document Classification using File Names
ABSTRACT: Rapid document classification is critical in several time-sensitive
applications like digital forensics and large-scale media classification.
Traditional approaches that rely on heavy-duty deep learning models fall short
due to high inference times over vast input datasets and computational
resources associated with analyzing whole documents. In this paper, we present
a method using lightweight supervised learning models, combined with a TF-IDF
feature extraction-based tokenization method, to accurately and efficiently
classify documents based solely on file names, that substantially reduces
inference time. Our results indicate that file name classifiers can process
more than 90% of in-scope documents with 99.63% and 96.57% accuracy when tested
on two datasets, while being 442x faster than more complex models such as DiT.
Our method offers a crucial solution to efficiently process vast document
datasets in critical scenarios, enabling fast and more reliable document
classification.
| no_new_dataset | 0.949949 |
2410.01257 | Zhilin Wang | Zhilin Wang, Alexander Bukharin, Olivier Delalleau, Daniel Egert,
Gerald Shen, Jiaqi Zeng, Oleksii Kuchaiev, Yi Dong | HelpSteer2-Preference: Complementing Ratings with Preferences | Accepted to ICLR 2025; 28 pages, 3 figures | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reward models are critical for aligning models to follow instructions, and
are typically trained following one of two popular paradigms: Bradley-Terry
style or Regression style. However, there is a lack of evidence that either
approach is better than the other, when adequately matched for data. This is
primarily because these approaches require data collected in different (but
incompatible) formats, meaning that adequately matched data is not available in
existing public datasets. To tackle this problem, we release preference
annotations (designed for Bradley-Terry training) to complement existing
ratings (designed for Regression style training) in the HelpSteer2 dataset. To
improve data interpretability, preference annotations are accompanied with
human-written justifications. Using this data, we conduct the first
head-to-head comparison of Bradley-Terry and Regression models when adequately
matched for data. Based on insights derived from such a comparison, we propose
a novel approach to combine Bradley-Terry and Regression reward modeling. A
Llama-3.1-70B-Instruct model tuned with this approach scores 94.1 on
RewardBench, emerging top of more than 140 reward models as of 1 Oct 2024. This
reward model can then be used with REINFORCE algorithm (RLHF) to align an
Instruct model to reach 85.0 on Arena Hard, which is No. 1 as of 1 Oct 2024. We
open-source this dataset (CC-BY-4.0 license) at
https://huggingface.co/datasets/nvidia/HelpSteer2#preferences-new -- 1-oct-2024
and openly release the trained Reward and Instruct models at
https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Reward and
https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct
| [
{
"version": "v1",
"created": "Wed, 2 Oct 2024 06:05:52 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 12:13:14 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"Zhilin",
""
],
[
"Bukharin",
"Alexander",
""
],
[
"Delalleau",
"Olivier",
""
],
[
"Egert",
"Daniel",
""
],
[
"Shen",
"Gerald",
""
],
[
"Zeng",
"Jiaqi",
""
],
[
"Kuchaiev",
"Oleksii",
""
],
[
"Dong",
"Yi",
""
]
]
| TITLE: HelpSteer2-Preference: Complementing Ratings with Preferences
ABSTRACT: Reward models are critical for aligning models to follow instructions, and
are typically trained following one of two popular paradigms: Bradley-Terry
style or Regression style. However, there is a lack of evidence that either
approach is better than the other, when adequately matched for data. This is
primarily because these approaches require data collected in different (but
incompatible) formats, meaning that adequately matched data is not available in
existing public datasets. To tackle this problem, we release preference
annotations (designed for Bradley-Terry training) to complement existing
ratings (designed for Regression style training) in the HelpSteer2 dataset. To
improve data interpretability, preference annotations are accompanied with
human-written justifications. Using this data, we conduct the first
head-to-head comparison of Bradley-Terry and Regression models when adequately
matched for data. Based on insights derived from such a comparison, we propose
a novel approach to combine Bradley-Terry and Regression reward modeling. A
Llama-3.1-70B-Instruct model tuned with this approach scores 94.1 on
RewardBench, emerging top of more than 140 reward models as of 1 Oct 2024. This
reward model can then be used with REINFORCE algorithm (RLHF) to align an
Instruct model to reach 85.0 on Arena Hard, which is No. 1 as of 1 Oct 2024. We
open-source this dataset (CC-BY-4.0 license) at
https://huggingface.co/datasets/nvidia/HelpSteer2#preferences-new -- 1-oct-2024
and openly release the trained Reward and Instruct models at
https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Reward and
https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct
| no_new_dataset | 0.942823 |
2410.01469 | Kai Li | Mohan Xu, Kai Li, Guo Chen, Xiaolin Hu | TIGER: Time-frequency Interleaved Gain Extraction and Reconstruction for
Efficient Speech Separation | Accepted by ICLR 2025, demo page: https://cslikai.cn/TIGER/ | null | null | null | cs.SD cs.AI eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, much speech separation research has focused primarily on
improving model performance. However, for low-latency speech processing
systems, high efficiency is equally important. Therefore, we propose a speech
separation model with significantly reduced parameters and computational costs:
Time-frequency Interleaved Gain Extraction and Reconstruction network (TIGER).
TIGER leverages prior knowledge to divide frequency bands and compresses
frequency information. We employ a multi-scale selective attention module to
extract contextual features while introducing a full-frequency-frame attention
module to capture both temporal and frequency contextual information.
Additionally, to more realistically evaluate the performance of speech
separation models in complex acoustic environments, we introduce a dataset
called EchoSet. This dataset includes noise and more realistic reverberation
(e.g., considering object occlusions and material properties), with speech from
two speakers overlapping at random proportions. Experimental results showed
that models trained on EchoSet had better generalization ability than those
trained on other datasets compared to the data collected in the physical world,
which validated the practical value of the EchoSet. On EchoSet and real-world
data, TIGER significantly reduces the number of parameters by 94.3% and the
MACs by 95.3% while achieving performance surpassing the state-of-the-art
(SOTA) model TF-GridNet.
| [
{
"version": "v1",
"created": "Wed, 2 Oct 2024 12:21:06 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 04:03:53 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Xu",
"Mohan",
""
],
[
"Li",
"Kai",
""
],
[
"Chen",
"Guo",
""
],
[
"Hu",
"Xiaolin",
""
]
]
| TITLE: TIGER: Time-frequency Interleaved Gain Extraction and Reconstruction for
Efficient Speech Separation
ABSTRACT: In recent years, much speech separation research has focused primarily on
improving model performance. However, for low-latency speech processing
systems, high efficiency is equally important. Therefore, we propose a speech
separation model with significantly reduced parameters and computational costs:
Time-frequency Interleaved Gain Extraction and Reconstruction network (TIGER).
TIGER leverages prior knowledge to divide frequency bands and compresses
frequency information. We employ a multi-scale selective attention module to
extract contextual features while introducing a full-frequency-frame attention
module to capture both temporal and frequency contextual information.
Additionally, to more realistically evaluate the performance of speech
separation models in complex acoustic environments, we introduce a dataset
called EchoSet. This dataset includes noise and more realistic reverberation
(e.g., considering object occlusions and material properties), with speech from
two speakers overlapping at random proportions. Experimental results showed
that models trained on EchoSet had better generalization ability than those
trained on other datasets compared to the data collected in the physical world,
which validated the practical value of the EchoSet. On EchoSet and real-world
data, TIGER significantly reduces the number of parameters by 94.3% and the
MACs by 95.3% while achieving performance surpassing the state-of-the-art
(SOTA) model TF-GridNet.
| new_dataset | 0.96502 |
2410.01481 | Kai Li | Kai Li, Wendi Sang, Chang Zeng, Runxuan Yang, Guo Chen, Xiaolin Hu | SonicSim: A customizable simulation platform for speech processing in
moving sound source scenarios | Accepted by ICLR 2025 | null | null | null | cs.SD cs.AI eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Systematic evaluation of speech separation and enhancement models under
moving sound source conditions requires extensive and diverse data. However,
real-world datasets often lack sufficient data for training and evaluation, and
synthetic datasets, while larger, lack acoustic realism. Consequently, neither
effectively meets practical needs. To address this issue, we introduce
SonicSim, a synthetic toolkit based on the embodied AI simulation platform
Habitat-sim, designed to generate highly customizable data for moving sound
sources. SonicSim supports multi-level adjustments, including scene-level,
microphone-level, and source-level adjustments, enabling the creation of more
diverse synthetic data. Leveraging SonicSim, we constructed a benchmark dataset
called SonicSet, utilizing LibriSpeech, Freesound Dataset 50k (FSD50K), Free
Music Archive (FMA), and 90 scenes from Matterport3D to evaluate speech
separation and enhancement models. Additionally, to investigate the differences
between synthetic and real-world data, we selected 5 hours of raw,
non-reverberant data from the SonicSet validation set and recorded a real-world
speech separation dataset, providing a reference for comparing SonicSet with
other synthetic datasets. For speech enhancement, we utilized the real-world
dataset RealMAN to validate the acoustic gap between SonicSet and existing
synthetic datasets. The results indicate that models trained on SonicSet
generalize better to real-world scenarios compared to other synthetic datasets.
The code is publicly available at https://cslikai.cn/SonicSim/.
| [
{
"version": "v1",
"created": "Wed, 2 Oct 2024 12:33:59 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 04:11:26 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Li",
"Kai",
""
],
[
"Sang",
"Wendi",
""
],
[
"Zeng",
"Chang",
""
],
[
"Yang",
"Runxuan",
""
],
[
"Chen",
"Guo",
""
],
[
"Hu",
"Xiaolin",
""
]
]
| TITLE: SonicSim: A customizable simulation platform for speech processing in
moving sound source scenarios
ABSTRACT: Systematic evaluation of speech separation and enhancement models under
moving sound source conditions requires extensive and diverse data. However,
real-world datasets often lack sufficient data for training and evaluation, and
synthetic datasets, while larger, lack acoustic realism. Consequently, neither
effectively meets practical needs. To address this issue, we introduce
SonicSim, a synthetic toolkit based on the embodied AI simulation platform
Habitat-sim, designed to generate highly customizable data for moving sound
sources. SonicSim supports multi-level adjustments, including scene-level,
microphone-level, and source-level adjustments, enabling the creation of more
diverse synthetic data. Leveraging SonicSim, we constructed a benchmark dataset
called SonicSet, utilizing LibriSpeech, Freesound Dataset 50k (FSD50K), Free
Music Archive (FMA), and 90 scenes from Matterport3D to evaluate speech
separation and enhancement models. Additionally, to investigate the differences
between synthetic and real-world data, we selected 5 hours of raw,
non-reverberant data from the SonicSet validation set and recorded a real-world
speech separation dataset, providing a reference for comparing SonicSet with
other synthetic datasets. For speech enhancement, we utilized the real-world
dataset RealMAN to validate the acoustic gap between SonicSet and existing
synthetic datasets. The results indicate that models trained on SonicSet
generalize better to real-world scenarios compared to other synthetic datasets.
The code is publicly available at https://cslikai.cn/SonicSim/.
| new_dataset | 0.964954 |
2410.06209 | Adarsh Kumarappan | Adarsh Kumarappan, Mo Tiwari, Peiyang Song, Robert Joseph George,
Chaowei Xiao, Anima Anandkumar | LeanAgent: Lifelong Learning for Formal Theorem Proving | null | null | null | null | cs.LG cs.AI cs.LO | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have been successful in mathematical reasoning
tasks such as formal theorem proving when integrated with interactive proof
assistants like Lean. Existing approaches involve training or fine-tuning an
LLM on a specific dataset to perform well on particular domains, such as
undergraduate-level mathematics. These methods struggle with generalizability
to advanced mathematics. A fundamental limitation is that these approaches
operate on static domains, failing to capture how mathematicians often work
across multiple domains and projects simultaneously or cyclically. We present
LeanAgent, a novel lifelong learning framework for formal theorem proving that
continuously generalizes to and improves on ever-expanding mathematical
knowledge without forgetting previously learned knowledge. LeanAgent introduces
several key innovations, including a curriculum learning strategy that
optimizes the learning trajectory in terms of mathematical difficulty, a
dynamic database for efficient management of evolving mathematical knowledge,
and progressive training to balance stability and plasticity. LeanAgent
successfully generates formal proofs for 155 theorems across 23 diverse Lean
repositories where formal proofs were previously missing, many from advanced
mathematics. It performs significantly better than the static LLM baseline,
proving challenging theorems in domains like abstract algebra and algebraic
topology while showcasing a clear progression of learning from basic concepts
to advanced topics. In addition, we analyze LeanAgent's superior performance on
key lifelong learning metrics. LeanAgent achieves exceptional scores in
stability and backward transfer, where learning new tasks improves performance
on previously learned tasks. This emphasizes LeanAgent's continuous
generalizability and improvement, explaining its superior theorem-proving
performance.
| [
{
"version": "v1",
"created": "Tue, 8 Oct 2024 17:11:24 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Oct 2024 05:20:07 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Oct 2024 19:26:56 GMT"
},
{
"version": "v4",
"created": "Fri, 18 Oct 2024 03:59:57 GMT"
},
{
"version": "v5",
"created": "Wed, 30 Oct 2024 05:20:25 GMT"
},
{
"version": "v6",
"created": "Wed, 6 Nov 2024 02:35:30 GMT"
},
{
"version": "v7",
"created": "Mon, 25 Nov 2024 02:39:03 GMT"
},
{
"version": "v8",
"created": "Thu, 6 Mar 2025 00:20:32 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Kumarappan",
"Adarsh",
""
],
[
"Tiwari",
"Mo",
""
],
[
"Song",
"Peiyang",
""
],
[
"George",
"Robert Joseph",
""
],
[
"Xiao",
"Chaowei",
""
],
[
"Anandkumar",
"Anima",
""
]
]
| TITLE: LeanAgent: Lifelong Learning for Formal Theorem Proving
ABSTRACT: Large Language Models (LLMs) have been successful in mathematical reasoning
tasks such as formal theorem proving when integrated with interactive proof
assistants like Lean. Existing approaches involve training or fine-tuning an
LLM on a specific dataset to perform well on particular domains, such as
undergraduate-level mathematics. These methods struggle with generalizability
to advanced mathematics. A fundamental limitation is that these approaches
operate on static domains, failing to capture how mathematicians often work
across multiple domains and projects simultaneously or cyclically. We present
LeanAgent, a novel lifelong learning framework for formal theorem proving that
continuously generalizes to and improves on ever-expanding mathematical
knowledge without forgetting previously learned knowledge. LeanAgent introduces
several key innovations, including a curriculum learning strategy that
optimizes the learning trajectory in terms of mathematical difficulty, a
dynamic database for efficient management of evolving mathematical knowledge,
and progressive training to balance stability and plasticity. LeanAgent
successfully generates formal proofs for 155 theorems across 23 diverse Lean
repositories where formal proofs were previously missing, many from advanced
mathematics. It performs significantly better than the static LLM baseline,
proving challenging theorems in domains like abstract algebra and algebraic
topology while showcasing a clear progression of learning from basic concepts
to advanced topics. In addition, we analyze LeanAgent's superior performance on
key lifelong learning metrics. LeanAgent achieves exceptional scores in
stability and backward transfer, where learning new tasks improves performance
on previously learned tasks. This emphasizes LeanAgent's continuous
generalizability and improvement, explaining its superior theorem-proving
performance.
| no_new_dataset | 0.947817 |
2410.10877 | Jinlong Pang | Jinlong Pang, Jiaheng Wei, Ankit Parag Shah, Zhaowei Zhu, Yaxuan Wang,
Chen Qian, Yang Liu, Yujia Bao, Wei Wei | Improving Data Efficiency via Curating LLM-Driven Rating Systems | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Instruction tuning is critical for adapting large language models (LLMs) to
downstream tasks, and recent studies have demonstrated that small amounts of
human-curated data can outperform larger datasets, challenging traditional data
scaling laws. While LLM-based data quality rating systems offer a
cost-effective alternative to human annotation, they often suffer from
inaccuracies and biases, even in powerful models like GPT-4. In this work, we
introduce DS2, a Diversity-aware Score curation method for Data Selection. By
systematically modeling error patterns through a score transition matrix, DS2
corrects LLM-based scores and promotes diversity in the selected data samples.
Our approach shows that a curated subset (just 3.3% of the original dataset)
outperforms full-scale datasets (300k samples) across various machine-alignment
benchmarks, and matches or surpasses human-aligned datasets such as LIMA with
the same sample size (1k samples). These findings challenge conventional data
scaling assumptions, highlighting that redundant, low-quality samples can
degrade performance and reaffirming that "more can be less."
| [
{
"version": "v1",
"created": "Wed, 9 Oct 2024 10:07:55 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 23:56:10 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Pang",
"Jinlong",
""
],
[
"Wei",
"Jiaheng",
""
],
[
"Shah",
"Ankit Parag",
""
],
[
"Zhu",
"Zhaowei",
""
],
[
"Wang",
"Yaxuan",
""
],
[
"Qian",
"Chen",
""
],
[
"Liu",
"Yang",
""
],
[
"Bao",
"Yujia",
""
],
[
"Wei",
"Wei",
""
]
]
| TITLE: Improving Data Efficiency via Curating LLM-Driven Rating Systems
ABSTRACT: Instruction tuning is critical for adapting large language models (LLMs) to
downstream tasks, and recent studies have demonstrated that small amounts of
human-curated data can outperform larger datasets, challenging traditional data
scaling laws. While LLM-based data quality rating systems offer a
cost-effective alternative to human annotation, they often suffer from
inaccuracies and biases, even in powerful models like GPT-4. In this work, we
introduce DS2, a Diversity-aware Score curation method for Data Selection. By
systematically modeling error patterns through a score transition matrix, DS2
corrects LLM-based scores and promotes diversity in the selected data samples.
Our approach shows that a curated subset (just 3.3% of the original dataset)
outperforms full-scale datasets (300k samples) across various machine-alignment
benchmarks, and matches or surpasses human-aligned datasets such as LIMA with
the same sample size (1k samples). These findings challenge conventional data
scaling assumptions, highlighting that redundant, low-quality samples can
degrade performance and reaffirming that "more can be less."
| no_new_dataset | 0.947039 |
2410.12261 | Xingjian Wu | Xingjian Wu, Xiangfei Qiu, Zhengyu Li, Yihang Wang, Jilin Hu, Chenjuan
Guo, Hui Xiong, Bin Yang | CATCH: Channel-Aware multivariate Time Series Anomaly Detection via
Frequency Patching | Accepted by ICLR 2025 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anomaly detection in multivariate time series is challenging as heterogeneous
subsequence anomalies may occur. Reconstruction-based methods, which focus on
learning normal patterns in the frequency domain to detect diverse abnormal
subsequences, achieve promising results, while still falling short on capturing
fine-grained frequency characteristics and channel correlations. To contend
with the limitations, we introduce CATCH, a framework based on frequency
patching. We propose to patchify the frequency domain into frequency bands,
which enhances its ability to capture fine-grained frequency characteristics.
To perceive appropriate channel correlations, we propose a Channel Fusion
Module (CFM), which features a patch-wise mask generator and a masked-attention
mechanism. Driven by a bi-level multi-objective optimization algorithm, the CFM
is encouraged to iteratively discover appropriate patch-wise channel
correlations, and to cluster relevant channels while isolating adverse effects
from irrelevant channels. Extensive experiments on 10 real-world datasets and
12 synthetic datasets demonstrate that CATCH achieves state-of-the-art
performance. We make our code and datasets available at
https://github.com/decisionintelligence/CATCH.
| [
{
"version": "v1",
"created": "Wed, 16 Oct 2024 05:58:55 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Feb 2025 06:59:46 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 13:39:32 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wu",
"Xingjian",
""
],
[
"Qiu",
"Xiangfei",
""
],
[
"Li",
"Zhengyu",
""
],
[
"Wang",
"Yihang",
""
],
[
"Hu",
"Jilin",
""
],
[
"Guo",
"Chenjuan",
""
],
[
"Xiong",
"Hui",
""
],
[
"Yang",
"Bin",
""
]
]
| TITLE: CATCH: Channel-Aware multivariate Time Series Anomaly Detection via
Frequency Patching
ABSTRACT: Anomaly detection in multivariate time series is challenging as heterogeneous
subsequence anomalies may occur. Reconstruction-based methods, which focus on
learning normal patterns in the frequency domain to detect diverse abnormal
subsequences, achieve promising results, while still falling short on capturing
fine-grained frequency characteristics and channel correlations. To contend
with the limitations, we introduce CATCH, a framework based on frequency
patching. We propose to patchify the frequency domain into frequency bands,
which enhances its ability to capture fine-grained frequency characteristics.
To perceive appropriate channel correlations, we propose a Channel Fusion
Module (CFM), which features a patch-wise mask generator and a masked-attention
mechanism. Driven by a bi-level multi-objective optimization algorithm, the CFM
is encouraged to iteratively discover appropriate patch-wise channel
correlations, and to cluster relevant channels while isolating adverse effects
from irrelevant channels. Extensive experiments on 10 real-world datasets and
12 synthetic datasets demonstrate that CATCH achieves state-of-the-art
performance. We make our code and datasets available at
https://github.com/decisionintelligence/CATCH.
| no_new_dataset | 0.94743 |
2410.14595 | Gao Yu Lee Mr. | Gao Yu Lee, Tanmoy Dam, Md Meftahul Ferdaus, Daniel Puiu Poenar, Vu
Duong | DRACO-DehazeNet: An Efficient Image Dehazing Network Combining Detail
Recovery and a Novel Contrastive Learning Paradigm | Once the paper is accepted and published, the copyright will be
transferred to the corresponding journal | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image dehazing is crucial for clarifying images obscured by haze or fog, but
current learning-based approaches is dependent on large volumes of training
data and hence consumed significant computational power. Additionally, their
performance is often inadequate under non-uniform or heavy haze. To address
these challenges, we developed the Detail Recovery And Contrastive DehazeNet,
which facilitates efficient and effective dehazing via a dense dilated inverted
residual block and an attention-based detail recovery network that tailors
enhancements to specific dehazed scene contexts. A major innovation is its
ability to train effectively with limited data, achieved through a novel
quadruplet loss-based contrastive dehazing paradigm. This approach distinctly
separates hazy and clear image features while also distinguish lower-quality
and higher-quality dehazed images obtained from each sub-modules of our
network, thereby refining the dehazing process to a larger extent. Extensive
tests on a variety of benchmarked haze datasets demonstrated the superiority of
our approach. The code repository for this work is available at
https://github.com/GreedYLearner1146/DRACO-DehazeNet.
| [
{
"version": "v1",
"created": "Fri, 18 Oct 2024 16:48:31 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 07:06:50 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Lee",
"Gao Yu",
""
],
[
"Dam",
"Tanmoy",
""
],
[
"Ferdaus",
"Md Meftahul",
""
],
[
"Poenar",
"Daniel Puiu",
""
],
[
"Duong",
"Vu",
""
]
]
| TITLE: DRACO-DehazeNet: An Efficient Image Dehazing Network Combining Detail
Recovery and a Novel Contrastive Learning Paradigm
ABSTRACT: Image dehazing is crucial for clarifying images obscured by haze or fog, but
current learning-based approaches is dependent on large volumes of training
data and hence consumed significant computational power. Additionally, their
performance is often inadequate under non-uniform or heavy haze. To address
these challenges, we developed the Detail Recovery And Contrastive DehazeNet,
which facilitates efficient and effective dehazing via a dense dilated inverted
residual block and an attention-based detail recovery network that tailors
enhancements to specific dehazed scene contexts. A major innovation is its
ability to train effectively with limited data, achieved through a novel
quadruplet loss-based contrastive dehazing paradigm. This approach distinctly
separates hazy and clear image features while also distinguish lower-quality
and higher-quality dehazed images obtained from each sub-modules of our
network, thereby refining the dehazing process to a larger extent. Extensive
tests on a variety of benchmarked haze datasets demonstrated the superiority of
our approach. The code repository for this work is available at
https://github.com/GreedYLearner1146/DRACO-DehazeNet.
| no_new_dataset | 0.943712 |
2410.17494 | Jun-En Ding | Jun-En Ding, Chien-Chin Hsu, Chi-Hsiang Chu, Shuqiang Wang, and Feng
Liu | Enhancing Multimodal Medical Image Classification using Cross-Graph
Modal Contrastive Learning | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | The classification of medical images is a pivotal aspect of disease
diagnosis, often enhanced by deep learning techniques. However, traditional
approaches typically focus on unimodal medical image data, neglecting the
integration of diverse non-image patient data. This paper proposes a novel
Cross-Graph Modal Contrastive Learning (CGMCL) framework for multimodal
structured data from different data domains to improve medical image
classification. The model effectively integrates both image and non-image data
by constructing cross-modality graphs and leveraging contrastive learning to
align multimodal features in a shared latent space. An inter-modality feature
scaling module further optimizes the representation learning process by
reducing the gap between heterogeneous modalities. The proposed approach is
evaluated on two datasets: a Parkinson's disease (PD) dataset and a public
melanoma dataset. Results demonstrate that CGMCL outperforms conventional
unimodal methods in accuracy, interpretability, and early disease prediction.
Additionally, the method shows superior performance in multi-class melanoma
classification. The CGMCL framework provides valuable insights into medical
image classification while offering improved disease interpretability and
predictive capabilities.
| [
{
"version": "v1",
"created": "Wed, 23 Oct 2024 01:25:25 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Nov 2024 18:35:14 GMT"
},
{
"version": "v3",
"created": "Fri, 7 Feb 2025 05:16:45 GMT"
},
{
"version": "v4",
"created": "Thu, 6 Mar 2025 16:43:10 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Ding",
"Jun-En",
""
],
[
"Hsu",
"Chien-Chin",
""
],
[
"Chu",
"Chi-Hsiang",
""
],
[
"Wang",
"Shuqiang",
""
],
[
"Liu",
"Feng",
""
]
]
| TITLE: Enhancing Multimodal Medical Image Classification using Cross-Graph
Modal Contrastive Learning
ABSTRACT: The classification of medical images is a pivotal aspect of disease
diagnosis, often enhanced by deep learning techniques. However, traditional
approaches typically focus on unimodal medical image data, neglecting the
integration of diverse non-image patient data. This paper proposes a novel
Cross-Graph Modal Contrastive Learning (CGMCL) framework for multimodal
structured data from different data domains to improve medical image
classification. The model effectively integrates both image and non-image data
by constructing cross-modality graphs and leveraging contrastive learning to
align multimodal features in a shared latent space. An inter-modality feature
scaling module further optimizes the representation learning process by
reducing the gap between heterogeneous modalities. The proposed approach is
evaluated on two datasets: a Parkinson's disease (PD) dataset and a public
melanoma dataset. Results demonstrate that CGMCL outperforms conventional
unimodal methods in accuracy, interpretability, and early disease prediction.
Additionally, the method shows superior performance in multi-class melanoma
classification. The CGMCL framework provides valuable insights into medical
image classification while offering improved disease interpretability and
predictive capabilities.
| no_new_dataset | 0.943815 |
2410.17635 | Wen Yang | Wen Yang, Minpeng Liao, Kai Fan | Markov Chain of Thought for Efficient Mathematical Reasoning | Camera ready version for NAACL 2025 Main | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Chain of Thought (CoT) of multi-step benefits from the logical structure of
the reasoning steps and task-specific actions, significantly enhancing the
mathematical reasoning capabilities of large language models. As the prevalence
of long CoT, the number of reasoning steps exceeds manageable token limits and
leads to higher computational demands. Inspired by the fundamental logic of
human cognition, "derive, then reduce", we conceptualize the standard
multi-step CoT as a novel Markov Chain of Thought (MCoT). In this study, we
consider the mathematical reasoning task, defining each reasoning step as text
accompanied by a Python code snippet. To facilitate a longer reasoning path,
self-correction is enabled through interactions with the code interpreter. Our
MCoT aims to compress previous reasoning steps into a simplified question,
enabling efficient next-step inference without relying on a lengthy KV cache.
In our experiments, we curate the $\texttt{MCoTInstruct}$ dataset, and the
empirical results indicate that MCoT not only significantly enhances efficiency
but also maintains comparable accuracy. While much remains to be explored, this
work paves the way for exploring the long CoT reasoning abilities of LLMs. The
code is available at https://github.com/james-yw/Markov-Chain-of-Thought
| [
{
"version": "v1",
"created": "Wed, 23 Oct 2024 07:53:29 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 06:39:56 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Yang",
"Wen",
""
],
[
"Liao",
"Minpeng",
""
],
[
"Fan",
"Kai",
""
]
]
| TITLE: Markov Chain of Thought for Efficient Mathematical Reasoning
ABSTRACT: Chain of Thought (CoT) of multi-step benefits from the logical structure of
the reasoning steps and task-specific actions, significantly enhancing the
mathematical reasoning capabilities of large language models. As the prevalence
of long CoT, the number of reasoning steps exceeds manageable token limits and
leads to higher computational demands. Inspired by the fundamental logic of
human cognition, "derive, then reduce", we conceptualize the standard
multi-step CoT as a novel Markov Chain of Thought (MCoT). In this study, we
consider the mathematical reasoning task, defining each reasoning step as text
accompanied by a Python code snippet. To facilitate a longer reasoning path,
self-correction is enabled through interactions with the code interpreter. Our
MCoT aims to compress previous reasoning steps into a simplified question,
enabling efficient next-step inference without relying on a lengthy KV cache.
In our experiments, we curate the $\texttt{MCoTInstruct}$ dataset, and the
empirical results indicate that MCoT not only significantly enhances efficiency
but also maintains comparable accuracy. While much remains to be explored, this
work paves the way for exploring the long CoT reasoning abilities of LLMs. The
code is available at https://github.com/james-yw/Markov-Chain-of-Thought
| new_dataset | 0.957118 |
2410.18653 | Esteban Garces Arias | Esteban Garces Arias and Hannah Blocher and Julian Rodemann and
Meimingwei Li and Christian Heumann and Matthias A{\ss}enmacher | Towards Better Open-Ended Text Generation: A Multicriteria Evaluation
Framework | null | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Open-ended text generation has become a prominent task in natural language
processing due to the rise of powerful (large) language models. However,
evaluating the quality of these models and the employed decoding strategies
remains challenging because of trade-offs among widely used metrics such as
coherence, diversity, and perplexity. Decoding methods often excel in some
metrics while underperforming in others, complicating the establishment of a
clear ranking. In this paper, we present novel ranking strategies within this
multicriteria framework. Specifically, we employ benchmarking approaches based
on partial orderings and present a new summary metric designed to balance
existing automatic indicators, providing a more holistic evaluation of text
generation quality. Our experiments demonstrate that the proposed methods offer
a robust way to compare decoding strategies, and serve as valuable tools in
guiding model selection for open-ended text generation tasks. Finally, we
suggest future directions for improving evaluation methodologies in text
generation. Our codebase, datasets, and models are publicly available.
| [
{
"version": "v1",
"created": "Thu, 24 Oct 2024 11:32:01 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 21:24:29 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Arias",
"Esteban Garces",
""
],
[
"Blocher",
"Hannah",
""
],
[
"Rodemann",
"Julian",
""
],
[
"Li",
"Meimingwei",
""
],
[
"Heumann",
"Christian",
""
],
[
"Aßenmacher",
"Matthias",
""
]
]
| TITLE: Towards Better Open-Ended Text Generation: A Multicriteria Evaluation
Framework
ABSTRACT: Open-ended text generation has become a prominent task in natural language
processing due to the rise of powerful (large) language models. However,
evaluating the quality of these models and the employed decoding strategies
remains challenging because of trade-offs among widely used metrics such as
coherence, diversity, and perplexity. Decoding methods often excel in some
metrics while underperforming in others, complicating the establishment of a
clear ranking. In this paper, we present novel ranking strategies within this
multicriteria framework. Specifically, we employ benchmarking approaches based
on partial orderings and present a new summary metric designed to balance
existing automatic indicators, providing a more holistic evaluation of text
generation quality. Our experiments demonstrate that the proposed methods offer
a robust way to compare decoding strategies, and serve as valuable tools in
guiding model selection for open-ended text generation tasks. Finally, we
suggest future directions for improving evaluation methodologies in text
generation. Our codebase, datasets, and models are publicly available.
| no_new_dataset | 0.945801 |
2410.24185 | Zhenyu Jiang | Zhenyu Jiang, Yuqi Xie, Kevin Lin, Zhenjia Xu, Weikang Wan, Ajay
Mandlekar, Linxi Fan, Yuke Zhu | DexMimicGen: Automated Data Generation for Bimanual Dexterous
Manipulation via Imitation Learning | ICRA 2025. Project website: https://dexmimicgen.github.io/ | null | null | null | cs.RO cs.AI cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Imitation learning from human demonstrations is an effective means to teach
robots manipulation skills. But data acquisition is a major bottleneck in
applying this paradigm more broadly, due to the amount of cost and human effort
involved. There has been significant interest in imitation learning for
bimanual dexterous robots, like humanoids. Unfortunately, data collection is
even more challenging here due to the challenges of simultaneously controlling
multiple arms and multi-fingered hands. Automated data generation in simulation
is a compelling, scalable alternative to fuel this need for data. To this end,
we introduce DexMimicGen, a large-scale automated data generation system that
synthesizes trajectories from a handful of human demonstrations for humanoid
robots with dexterous hands. We present a collection of simulation environments
in the setting of bimanual dexterous manipulation, spanning a range of
manipulation behaviors and different requirements for coordination among the
two arms. We generate 21K demos across these tasks from just 60 source human
demos and study the effect of several data generation and policy learning
decisions on agent performance. Finally, we present a real-to-sim-to-real
pipeline and deploy it on a real-world humanoid can sorting task. Generated
datasets, simulation environments and additional results are at
https://dexmimicgen.github.io/
| [
{
"version": "v1",
"created": "Thu, 31 Oct 2024 17:48:45 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 05:34:17 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Jiang",
"Zhenyu",
""
],
[
"Xie",
"Yuqi",
""
],
[
"Lin",
"Kevin",
""
],
[
"Xu",
"Zhenjia",
""
],
[
"Wan",
"Weikang",
""
],
[
"Mandlekar",
"Ajay",
""
],
[
"Fan",
"Linxi",
""
],
[
"Zhu",
"Yuke",
""
]
]
| TITLE: DexMimicGen: Automated Data Generation for Bimanual Dexterous
Manipulation via Imitation Learning
ABSTRACT: Imitation learning from human demonstrations is an effective means to teach
robots manipulation skills. But data acquisition is a major bottleneck in
applying this paradigm more broadly, due to the amount of cost and human effort
involved. There has been significant interest in imitation learning for
bimanual dexterous robots, like humanoids. Unfortunately, data collection is
even more challenging here due to the challenges of simultaneously controlling
multiple arms and multi-fingered hands. Automated data generation in simulation
is a compelling, scalable alternative to fuel this need for data. To this end,
we introduce DexMimicGen, a large-scale automated data generation system that
synthesizes trajectories from a handful of human demonstrations for humanoid
robots with dexterous hands. We present a collection of simulation environments
in the setting of bimanual dexterous manipulation, spanning a range of
manipulation behaviors and different requirements for coordination among the
two arms. We generate 21K demos across these tasks from just 60 source human
demos and study the effect of several data generation and policy learning
decisions on agent performance. Finally, we present a real-to-sim-to-real
pipeline and deploy it on a real-world humanoid can sorting task. Generated
datasets, simulation environments and additional results are at
https://dexmimicgen.github.io/
| no_new_dataset | 0.746786 |
2411.07076 | Yichen He | Yichen He, Yuan Lin, Jianchao Wu, Hanchong Zhang, Yuchen Zhang,
Ruicheng Le | StoryTeller: Improving Long Video Description through Global
Audio-Visual Character Identification | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Existing large vision-language models (LVLMs) are largely limited to
processing short, seconds-long videos and struggle with generating coherent
descriptions for extended video spanning minutes or more. Long video
description introduces new challenges, such as consistent character
identification and plot-level descriptions incorporating both visual and audio
information. To address these, we figure out audio-visual character
identification, matching character names to each dialogue, as a key factor. We
propose StoryTeller, a system for generating dense descriptions of long videos,
incorporating both low-level visual concepts and high-level plot information.
StoryTeller uses a multimodal large language model that integrates visual,
audio, and text modalities to perform audio-visual character identification on
minute-long video clips. The results are then fed into a LVLM to enhance
consistency of video description. We validate our approach on movie description
tasks and introduce MovieStory101, a dataset with dense descriptions for
three-minute movie clips. To evaluate long video descriptions, we create
StoryQA, a large set of multiple-choice questions for MovieStory101 test set.
We assess descriptions by inputting them into GPT-4 to answer these questions,
using accuracy as an automatic evaluation metric. Experiments show that
StoryTeller outperforms all open and closed-source baselines on StoryQA,
achieving 9.5% higher accuracy than the strongest baseline, Gemini-1.5-pro, and
demonstrating a +15.56% advantage in human side-by-side evaluations.
Additionally, incorporating audio-visual character identification from
StoryTeller improves the performance of all video description models, with
Gemini-1.5-pro and GPT-4o showing relative improvement of 5.5% and 13.0%,
respectively, in accuracy on StoryQA.
| [
{
"version": "v1",
"created": "Mon, 11 Nov 2024 15:51:48 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 09:13:28 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"He",
"Yichen",
""
],
[
"Lin",
"Yuan",
""
],
[
"Wu",
"Jianchao",
""
],
[
"Zhang",
"Hanchong",
""
],
[
"Zhang",
"Yuchen",
""
],
[
"Le",
"Ruicheng",
""
]
]
| TITLE: StoryTeller: Improving Long Video Description through Global
Audio-Visual Character Identification
ABSTRACT: Existing large vision-language models (LVLMs) are largely limited to
processing short, seconds-long videos and struggle with generating coherent
descriptions for extended video spanning minutes or more. Long video
description introduces new challenges, such as consistent character
identification and plot-level descriptions incorporating both visual and audio
information. To address these, we figure out audio-visual character
identification, matching character names to each dialogue, as a key factor. We
propose StoryTeller, a system for generating dense descriptions of long videos,
incorporating both low-level visual concepts and high-level plot information.
StoryTeller uses a multimodal large language model that integrates visual,
audio, and text modalities to perform audio-visual character identification on
minute-long video clips. The results are then fed into a LVLM to enhance
consistency of video description. We validate our approach on movie description
tasks and introduce MovieStory101, a dataset with dense descriptions for
three-minute movie clips. To evaluate long video descriptions, we create
StoryQA, a large set of multiple-choice questions for MovieStory101 test set.
We assess descriptions by inputting them into GPT-4 to answer these questions,
using accuracy as an automatic evaluation metric. Experiments show that
StoryTeller outperforms all open and closed-source baselines on StoryQA,
achieving 9.5% higher accuracy than the strongest baseline, Gemini-1.5-pro, and
demonstrating a +15.56% advantage in human side-by-side evaluations.
Additionally, incorporating audio-visual character identification from
StoryTeller improves the performance of all video description models, with
Gemini-1.5-pro and GPT-4o showing relative improvement of 5.5% and 13.0%,
respectively, in accuracy on StoryQA.
| new_dataset | 0.932269 |
2411.08410 | Yangyang Guo | Yangyang Guo and Fangkai Jiao and Liqiang Nie and Mohan Kankanhalli | The VLLM Safety Paradox: Dual Ease in Jailbreak Attack and Defense | Logic smoothing and language polishing | null | null | null | cs.CR cs.CV | http://creativecommons.org/licenses/by/4.0/ | The vulnerability of Vision Large Language Models (VLLMs) to jailbreak
attacks appears as no surprise. However, recent defense mechanisms against
these attacks have reached near-saturation performance on benchmark
evaluations, often with minimal effort. This \emph{dual high performance} in
both attack and defense raises a fundamental and perplexing paradox. To gain a
deep understanding of this issue and thus further help strengthen the
trustworthiness of VLLMs, this paper makes three key contributions: i) One
tentative explanation for VLLMs being prone to jailbreak
attacks--\textbf{inclusion of vision inputs}, as well as its in-depth analysis.
ii) The recognition of a largely ignored problem in existing defense
mechanisms--\textbf{over-prudence}. The problem causes these defense methods to
exhibit unintended abstention, even in the presence of benign inputs, thereby
undermining their reliability in faithfully defending against attacks. iii) A
simple safety-aware method--\textbf{LLM-Pipeline}. Our method repurposes the
more advanced guardrails of LLMs on the shelf, serving as an effective
alternative detector prior to VLLM response. Last but not least, we find that
the two representative evaluation methods for jailbreak often exhibit chance
agreement. This limitation makes it potentially misleading when evaluating
attack strategies or defense mechanisms. We believe the findings from this
paper offer useful insights to rethink the foundational development of VLLM
safety with respect to benchmark datasets, defense strategies, and evaluation
methods.
| [
{
"version": "v1",
"created": "Wed, 13 Nov 2024 07:57:19 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 01:45:26 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Guo",
"Yangyang",
""
],
[
"Jiao",
"Fangkai",
""
],
[
"Nie",
"Liqiang",
""
],
[
"Kankanhalli",
"Mohan",
""
]
]
| TITLE: The VLLM Safety Paradox: Dual Ease in Jailbreak Attack and Defense
ABSTRACT: The vulnerability of Vision Large Language Models (VLLMs) to jailbreak
attacks appears as no surprise. However, recent defense mechanisms against
these attacks have reached near-saturation performance on benchmark
evaluations, often with minimal effort. This \emph{dual high performance} in
both attack and defense raises a fundamental and perplexing paradox. To gain a
deep understanding of this issue and thus further help strengthen the
trustworthiness of VLLMs, this paper makes three key contributions: i) One
tentative explanation for VLLMs being prone to jailbreak
attacks--\textbf{inclusion of vision inputs}, as well as its in-depth analysis.
ii) The recognition of a largely ignored problem in existing defense
mechanisms--\textbf{over-prudence}. The problem causes these defense methods to
exhibit unintended abstention, even in the presence of benign inputs, thereby
undermining their reliability in faithfully defending against attacks. iii) A
simple safety-aware method--\textbf{LLM-Pipeline}. Our method repurposes the
more advanced guardrails of LLMs on the shelf, serving as an effective
alternative detector prior to VLLM response. Last but not least, we find that
the two representative evaluation methods for jailbreak often exhibit chance
agreement. This limitation makes it potentially misleading when evaluating
attack strategies or defense mechanisms. We believe the findings from this
paper offer useful insights to rethink the foundational development of VLLM
safety with respect to benchmark datasets, defense strategies, and evaluation
methods.
| no_new_dataset | 0.945901 |
2411.09263 | Hu Wang | Hu Wang, Congbo Ma, Ibrahim Almakky, Ian Reid, Gustavo Carneiro,
Mohammad Yaqub | Rethinking Weight-Averaged Model-merging | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Model-merging has emerged as a powerful approach in deep learning, capable of
enhancing model performance without any training. However, the underlying
mechanisms that explain its effectiveness remain largely unexplored. In this
paper, we investigate this technique from three novel perspectives to
empirically provide deeper insights into why and how weight-averaged
model-merging works: (1) we examine the intrinsic patterns captured by the
learning of the model weights, through the visualizations of their patterns on
several datasets, showing that these weights often encode structured and
interpretable patterns and that is the essential why model-merging can work;
(2) we mathematically and empirically investigate model ensemble merging
strategies based on averaging on weights versus averaging on features,
providing detailed analyses across diverse architectures and datasets; and (3)
we explore the impact on model-merging prediction stability in terms of
changing the parameter magnitude, revealing insights into the way of weight
averaging works as regularization by showing the robustness across different
parameter scales. Our findings shed light on the "black box" of weight-averaged
model-merging, offering valuable insights and practical recommendations that
advance the model-merging process. The code is available at
https://github.com/billhhh/Rethink-Merge.
| [
{
"version": "v1",
"created": "Thu, 14 Nov 2024 08:02:14 GMT"
},
{
"version": "v2",
"created": "Thu, 21 Nov 2024 10:46:18 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 09:10:07 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"Hu",
""
],
[
"Ma",
"Congbo",
""
],
[
"Almakky",
"Ibrahim",
""
],
[
"Reid",
"Ian",
""
],
[
"Carneiro",
"Gustavo",
""
],
[
"Yaqub",
"Mohammad",
""
]
]
| TITLE: Rethinking Weight-Averaged Model-merging
ABSTRACT: Model-merging has emerged as a powerful approach in deep learning, capable of
enhancing model performance without any training. However, the underlying
mechanisms that explain its effectiveness remain largely unexplored. In this
paper, we investigate this technique from three novel perspectives to
empirically provide deeper insights into why and how weight-averaged
model-merging works: (1) we examine the intrinsic patterns captured by the
learning of the model weights, through the visualizations of their patterns on
several datasets, showing that these weights often encode structured and
interpretable patterns and that is the essential why model-merging can work;
(2) we mathematically and empirically investigate model ensemble merging
strategies based on averaging on weights versus averaging on features,
providing detailed analyses across diverse architectures and datasets; and (3)
we explore the impact on model-merging prediction stability in terms of
changing the parameter magnitude, revealing insights into the way of weight
averaging works as regularization by showing the robustness across different
parameter scales. Our findings shed light on the "black box" of weight-averaged
model-merging, offering valuable insights and practical recommendations that
advance the model-merging process. The code is available at
https://github.com/billhhh/Rethink-Merge.
| no_new_dataset | 0.941223 |
2411.11006 | Haiyang Yu | Haiyang Yu, Tian Xie, Jiaping Gui, Pengyang Wang, Ping Yi, Yue Wu | BackdoorMBTI: A Backdoor Learning Multimodal Benchmark Tool Kit for
Backdoor Defense Evaluation | null | null | 10.1145/3690624.3709385 | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Over the past few years, the emergence of backdoor attacks has presented
significant challenges to deep learning systems, allowing attackers to insert
backdoors into neural networks. When data with a trigger is processed by a
backdoor model, it can lead to mispredictions targeted by attackers, whereas
normal data yields regular results. The scope of backdoor attacks is expanding
beyond computer vision and encroaching into areas such as natural language
processing and speech recognition. Nevertheless, existing backdoor defense
methods are typically tailored to specific data modalities, restricting their
application in multimodal contexts. While multimodal learning proves highly
applicable in facial recognition, sentiment analysis, action recognition,
visual question answering, the security of these models remains a crucial
concern. Specifically, there are no existing backdoor benchmarks targeting
multimodal applications or related tasks.
In order to facilitate the research in multimodal backdoor, we introduce
BackdoorMBTI, the first backdoor learning toolkit and benchmark designed for
multimodal evaluation across three representative modalities from eleven
commonly used datasets. BackdoorMBTI provides a systematic backdoor learning
pipeline, encompassing data processing, data poisoning, backdoor training, and
evaluation. The generated poison datasets and backdoor models enable detailed
evaluation of backdoor defenses. Given the diversity of modalities,
BackdoorMBTI facilitates systematic evaluation across different data types.
Furthermore, BackdoorMBTI offers a standardized approach to handling practical
factors in backdoor learning, such as issues related to data quality and
erroneous labels. We anticipate that BackdoorMBTI will expedite future research
in backdoor defense methods within a multimodal context. Code is available at
https://github.com/SJTUHaiyangYu/BackdoorMBTI.
| [
{
"version": "v1",
"created": "Sun, 17 Nov 2024 09:01:55 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 07:50:21 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Yu",
"Haiyang",
""
],
[
"Xie",
"Tian",
""
],
[
"Gui",
"Jiaping",
""
],
[
"Wang",
"Pengyang",
""
],
[
"Yi",
"Ping",
""
],
[
"Wu",
"Yue",
""
]
]
| TITLE: BackdoorMBTI: A Backdoor Learning Multimodal Benchmark Tool Kit for
Backdoor Defense Evaluation
ABSTRACT: Over the past few years, the emergence of backdoor attacks has presented
significant challenges to deep learning systems, allowing attackers to insert
backdoors into neural networks. When data with a trigger is processed by a
backdoor model, it can lead to mispredictions targeted by attackers, whereas
normal data yields regular results. The scope of backdoor attacks is expanding
beyond computer vision and encroaching into areas such as natural language
processing and speech recognition. Nevertheless, existing backdoor defense
methods are typically tailored to specific data modalities, restricting their
application in multimodal contexts. While multimodal learning proves highly
applicable in facial recognition, sentiment analysis, action recognition,
visual question answering, the security of these models remains a crucial
concern. Specifically, there are no existing backdoor benchmarks targeting
multimodal applications or related tasks.
In order to facilitate the research in multimodal backdoor, we introduce
BackdoorMBTI, the first backdoor learning toolkit and benchmark designed for
multimodal evaluation across three representative modalities from eleven
commonly used datasets. BackdoorMBTI provides a systematic backdoor learning
pipeline, encompassing data processing, data poisoning, backdoor training, and
evaluation. The generated poison datasets and backdoor models enable detailed
evaluation of backdoor defenses. Given the diversity of modalities,
BackdoorMBTI facilitates systematic evaluation across different data types.
Furthermore, BackdoorMBTI offers a standardized approach to handling practical
factors in backdoor learning, such as issues related to data quality and
erroneous labels. We anticipate that BackdoorMBTI will expedite future research
in backdoor defense methods within a multimodal context. Code is available at
https://github.com/SJTUHaiyangYu/BackdoorMBTI.
| no_new_dataset | 0.937268 |
2411.11505 | Zhaoqing Wang | Zhaoqing Wang, Xiaobo Xia, Runnan Chen, Dongdong Yu, Changhu Wang,
Mingming Gong, Tongliang Liu | LaVin-DiT: Large Vision Diffusion Transformer | 37 pages, 30 figures, 4 tables. Accepted by CVPR 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper presents the Large Vision Diffusion Transformer (LaVin-DiT), a
scalable and unified foundation model designed to tackle over 20 computer
vision tasks in a generative framework. Unlike existing large vision models
directly adapted from natural language processing architectures, which rely on
less efficient autoregressive techniques and disrupt spatial relationships
essential for vision data, LaVin-DiT introduces key innovations to optimize
generative performance for vision tasks. First, to address the high
dimensionality of visual data, we incorporate a spatial-temporal variational
autoencoder that encodes data into a continuous latent space. Second, for
generative modeling, we develop a joint diffusion transformer that
progressively produces vision outputs. Third, for unified multi-task training,
in-context learning is implemented. Input-target pairs serve as task context,
which guides the diffusion transformer to align outputs with specific tasks
within the latent space. During inference, a task-specific context set and test
data as queries allow LaVin-DiT to generalize across tasks without fine-tuning.
Trained on extensive vision datasets, the model is scaled from 0.1B to 3.4B
parameters, demonstrating substantial scalability and state-of-the-art
performance across diverse vision tasks. This work introduces a novel pathway
for large vision foundation models, underscoring the promising potential of
diffusion transformers. The code and models are available.
| [
{
"version": "v1",
"created": "Mon, 18 Nov 2024 12:05:27 GMT"
},
{
"version": "v2",
"created": "Sat, 23 Nov 2024 21:10:24 GMT"
},
{
"version": "v3",
"created": "Tue, 26 Nov 2024 06:48:45 GMT"
},
{
"version": "v4",
"created": "Thu, 6 Mar 2025 07:26:35 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"Zhaoqing",
""
],
[
"Xia",
"Xiaobo",
""
],
[
"Chen",
"Runnan",
""
],
[
"Yu",
"Dongdong",
""
],
[
"Wang",
"Changhu",
""
],
[
"Gong",
"Mingming",
""
],
[
"Liu",
"Tongliang",
""
]
]
| TITLE: LaVin-DiT: Large Vision Diffusion Transformer
ABSTRACT: This paper presents the Large Vision Diffusion Transformer (LaVin-DiT), a
scalable and unified foundation model designed to tackle over 20 computer
vision tasks in a generative framework. Unlike existing large vision models
directly adapted from natural language processing architectures, which rely on
less efficient autoregressive techniques and disrupt spatial relationships
essential for vision data, LaVin-DiT introduces key innovations to optimize
generative performance for vision tasks. First, to address the high
dimensionality of visual data, we incorporate a spatial-temporal variational
autoencoder that encodes data into a continuous latent space. Second, for
generative modeling, we develop a joint diffusion transformer that
progressively produces vision outputs. Third, for unified multi-task training,
in-context learning is implemented. Input-target pairs serve as task context,
which guides the diffusion transformer to align outputs with specific tasks
within the latent space. During inference, a task-specific context set and test
data as queries allow LaVin-DiT to generalize across tasks without fine-tuning.
Trained on extensive vision datasets, the model is scaled from 0.1B to 3.4B
parameters, demonstrating substantial scalability and state-of-the-art
performance across diverse vision tasks. This work introduces a novel pathway
for large vision foundation models, underscoring the promising potential of
diffusion transformers. The code and models are available.
| no_new_dataset | 0.946349 |
2411.13056 | Quanhao Lu | Bing Cao, Quanhao Lu, Jiekang Feng, Qilong Wang, Qinghua Hu, Pengfei
Zhu | Efficient Masked AutoEncoder for Video Object Counting and A Large-Scale
Benchmark | ICLR25 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The dynamic imbalance of the fore-background is a major challenge in video
object counting, which is usually caused by the sparsity of target objects.
This remains understudied in existing works and often leads to severe
under-/over-prediction errors. To tackle this issue in video object counting,
we propose a density-embedded Efficient Masked Autoencoder Counting (E-MAC)
framework in this paper. To empower the model's representation ability on
density regression, we develop a new $\mathtt{D}$ensity-$\mathtt{E}$mbedded
$\mathtt{M}$asked m$\mathtt{O}$deling ($\mathtt{DEMO}$) method, which first
takes the density map as an auxiliary modality to perform multimodal
self-representation learning for image and density map. Although
$\mathtt{DEMO}$ contributes to effective cross-modal regression guidance, it
also brings in redundant background information, making it difficult to focus
on the foreground regions. To handle this dilemma, we propose an efficient
spatial adaptive masking derived from density maps to boost efficiency.
Meanwhile, we employ an optical flow-based temporal collaborative fusion
strategy to effectively capture the dynamic variations across frames, aligning
features to derive multi-frame density residuals. The counting accuracy of the
current frame is boosted by harnessing the information from adjacent frames. In
addition, considering that most existing datasets are limited to human-centric
scenarios, we first propose a large video bird counting dataset, DroneBird, in
natural scenarios for migratory bird protection. Extensive experiments on three
crowd datasets and our \textit{DroneBird} validate our superiority against the
counterparts. The code and dataset are available.
| [
{
"version": "v1",
"created": "Wed, 20 Nov 2024 06:08:21 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 08:28:09 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Cao",
"Bing",
""
],
[
"Lu",
"Quanhao",
""
],
[
"Feng",
"Jiekang",
""
],
[
"Wang",
"Qilong",
""
],
[
"Hu",
"Qinghua",
""
],
[
"Zhu",
"Pengfei",
""
]
]
| TITLE: Efficient Masked AutoEncoder for Video Object Counting and A Large-Scale
Benchmark
ABSTRACT: The dynamic imbalance of the fore-background is a major challenge in video
object counting, which is usually caused by the sparsity of target objects.
This remains understudied in existing works and often leads to severe
under-/over-prediction errors. To tackle this issue in video object counting,
we propose a density-embedded Efficient Masked Autoencoder Counting (E-MAC)
framework in this paper. To empower the model's representation ability on
density regression, we develop a new $\mathtt{D}$ensity-$\mathtt{E}$mbedded
$\mathtt{M}$asked m$\mathtt{O}$deling ($\mathtt{DEMO}$) method, which first
takes the density map as an auxiliary modality to perform multimodal
self-representation learning for image and density map. Although
$\mathtt{DEMO}$ contributes to effective cross-modal regression guidance, it
also brings in redundant background information, making it difficult to focus
on the foreground regions. To handle this dilemma, we propose an efficient
spatial adaptive masking derived from density maps to boost efficiency.
Meanwhile, we employ an optical flow-based temporal collaborative fusion
strategy to effectively capture the dynamic variations across frames, aligning
features to derive multi-frame density residuals. The counting accuracy of the
current frame is boosted by harnessing the information from adjacent frames. In
addition, considering that most existing datasets are limited to human-centric
scenarios, we first propose a large video bird counting dataset, DroneBird, in
natural scenarios for migratory bird protection. Extensive experiments on three
crowd datasets and our \textit{DroneBird} validate our superiority against the
counterparts. The code and dataset are available.
| new_dataset | 0.864024 |
2411.16157 | Chenjie Cao | Chenjie Cao, Chaohui Yu, Shang Liu, Fan Wang, Xiangyang Xue, Yanwei Fu | MVGenMaster: Scaling Multi-View Generation from Any Image via 3D Priors
Enhanced Diffusion Model | Accepted by CVPR2025. Models and codes will be released at
https://github.com/ewrfcas/MVGenMaster/. The project page is at
https://ewrfcas.github.io/MVGenMaster/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce MVGenMaster, a multi-view diffusion model enhanced with 3D
priors to address versatile Novel View Synthesis (NVS) tasks. MVGenMaster
leverages 3D priors that are warped using metric depth and camera poses,
significantly enhancing both generalization and 3D consistency in NVS. Our
model features a simple yet effective pipeline that can generate up to 100
novel views conditioned on variable reference views and camera poses with a
single forward process. Additionally, we have developed a comprehensive
large-scale multi-view image dataset called MvD-1M, comprising up to 1.6
million scenes, equipped with well-aligned metric depth to train MVGenMaster.
Moreover, we present several training and model modifications to strengthen the
model with scaled-up datasets. Extensive evaluations across in- and
out-of-domain benchmarks demonstrate the effectiveness of our proposed method
and data formulation. Models and codes will be released at
https://github.com/ewrfcas/MVGenMaster/.
| [
{
"version": "v1",
"created": "Mon, 25 Nov 2024 07:34:23 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Nov 2024 06:33:58 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 02:45:21 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Cao",
"Chenjie",
""
],
[
"Yu",
"Chaohui",
""
],
[
"Liu",
"Shang",
""
],
[
"Wang",
"Fan",
""
],
[
"Xue",
"Xiangyang",
""
],
[
"Fu",
"Yanwei",
""
]
]
| TITLE: MVGenMaster: Scaling Multi-View Generation from Any Image via 3D Priors
Enhanced Diffusion Model
ABSTRACT: We introduce MVGenMaster, a multi-view diffusion model enhanced with 3D
priors to address versatile Novel View Synthesis (NVS) tasks. MVGenMaster
leverages 3D priors that are warped using metric depth and camera poses,
significantly enhancing both generalization and 3D consistency in NVS. Our
model features a simple yet effective pipeline that can generate up to 100
novel views conditioned on variable reference views and camera poses with a
single forward process. Additionally, we have developed a comprehensive
large-scale multi-view image dataset called MvD-1M, comprising up to 1.6
million scenes, equipped with well-aligned metric depth to train MVGenMaster.
Moreover, we present several training and model modifications to strengthen the
model with scaled-up datasets. Extensive evaluations across in- and
out-of-domain benchmarks demonstrate the effectiveness of our proposed method
and data formulation. Models and codes will be released at
https://github.com/ewrfcas/MVGenMaster/.
| new_dataset | 0.953837 |
2412.00411 | Mohammad Hasan Rahmani | Mohammad Hasan Rahmani, Rafael Berkvens and Maarten Weyn | Seismocardiography for Emotion Recognition: A Study on EmoWear with
Insights from DEAP | 16 pages, 9 figures | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Emotions have a profound impact on our daily lives, influencing our thoughts,
behaviors, and interactions, but also our physiological reactions. Recent
advances in wearable technology have facilitated studying emotions through
cardio-respiratory signals. Accelerometers offer a non-invasive, convenient,
and cost-effective method for capturing heart- and pulmonary-induced vibrations
on the chest wall, specifically Seismocardiography (SCG) and
Accelerometry-Derived Respiration (ADR). Their affordability, wide
availability, and ability to provide rich contextual data make accelerometers
ideal for everyday use. While accelerometers have been used as part of broader
modality fusions for Emotion Recognition (ER), their stand-alone potential via
SCG and ADR remains unexplored. Bridging this gap could significantly help the
embedding of ER into real-world applications. To address this gap, we introduce
SCG as a novel modality for ER and evaluate its performance using the EmoWear
dataset. First, we replicate the single-trial emotion classification pipeline
from the DEAP dataset study, achieving similar results. Then we use our
validated pipeline to train models that predict affective valence-arousal
states using SCG and compare them against established cardiac signals,
Electrocardiography (ECG) and Blood Volume Pulse (BVP). Results show that SCG
is a viable modality for ER, achieving similar performance to ECG and BVP. By
combining ADR with SCG, we achieved a working ER framework that only requires a
single chest-worn accelerometer. These findings pave the way for integrating ER
into real-world, enabling seamless affective computing in everyday life.
| [
{
"version": "v1",
"created": "Sat, 30 Nov 2024 09:39:30 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 22:03:48 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Rahmani",
"Mohammad Hasan",
""
],
[
"Berkvens",
"Rafael",
""
],
[
"Weyn",
"Maarten",
""
]
]
| TITLE: Seismocardiography for Emotion Recognition: A Study on EmoWear with
Insights from DEAP
ABSTRACT: Emotions have a profound impact on our daily lives, influencing our thoughts,
behaviors, and interactions, but also our physiological reactions. Recent
advances in wearable technology have facilitated studying emotions through
cardio-respiratory signals. Accelerometers offer a non-invasive, convenient,
and cost-effective method for capturing heart- and pulmonary-induced vibrations
on the chest wall, specifically Seismocardiography (SCG) and
Accelerometry-Derived Respiration (ADR). Their affordability, wide
availability, and ability to provide rich contextual data make accelerometers
ideal for everyday use. While accelerometers have been used as part of broader
modality fusions for Emotion Recognition (ER), their stand-alone potential via
SCG and ADR remains unexplored. Bridging this gap could significantly help the
embedding of ER into real-world applications. To address this gap, we introduce
SCG as a novel modality for ER and evaluate its performance using the EmoWear
dataset. First, we replicate the single-trial emotion classification pipeline
from the DEAP dataset study, achieving similar results. Then we use our
validated pipeline to train models that predict affective valence-arousal
states using SCG and compare them against established cardiac signals,
Electrocardiography (ECG) and Blood Volume Pulse (BVP). Results show that SCG
is a viable modality for ER, achieving similar performance to ECG and BVP. By
combining ADR with SCG, we achieved a working ER framework that only requires a
single chest-worn accelerometer. These findings pave the way for integrating ER
into real-world, enabling seamless affective computing in everyday life.
| no_new_dataset | 0.943971 |
2412.07093 | Joel Daniel Andersson | Joel Daniel Andersson, Rasmus Pagh | Streaming Private Continual Counting via Binning | Accepted to SaTML 2025. Final version to appear on IEEE eXplore | null | null | null | cs.LG cs.CR cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In differential privacy, $\textit{continual observation}$ refers to problems
in which we wish to continuously release a function of a dataset that is
revealed one element at a time. The challenge is to maintain a good
approximation while keeping the combined output over all time steps
differentially private. In the special case of $\textit{continual counting}$ we
seek to approximate a sum of binary input elements. This problem has received
considerable attention lately, in part due to its relevance in implementations
of differentially private stochastic gradient descent. $\textit{Factorization
mechanisms}$ are the leading approach to continual counting, but the best such
mechanisms do not work well in $\textit{streaming}$ settings since they require
space proportional to the size of the input. In this paper, we present a simple
approach to approximating factorization mechanisms in low space via
$\textit{binning}$, where adjacent matrix entries with similar values are
changed to be identical in such a way that a matrix-vector product can be
maintained in sublinear space. Our approach has provable sublinear space
guarantees for a class of lower triangular matrices whose entries are
monotonically decreasing away from the diagonal. We show empirically that even
with very low space usage we are able to closely match, and sometimes surpass,
the performance of asymptotically optimal factorization mechanisms. Recently,
and independently of our work, Dvijotham et al. have also suggested an approach
to implementing factorization mechanisms in a streaming setting. Their work
differs from ours in several respects: It only addresses factorization into
$\textit{Toeplitz}$ matrices, only considers $\textit{maximum}$ error, and uses
a different technique based on rational function approximation that seems less
versatile than our binning approach.
| [
{
"version": "v1",
"created": "Tue, 10 Dec 2024 01:21:56 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 16:14:01 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Andersson",
"Joel Daniel",
""
],
[
"Pagh",
"Rasmus",
""
]
]
| TITLE: Streaming Private Continual Counting via Binning
ABSTRACT: In differential privacy, $\textit{continual observation}$ refers to problems
in which we wish to continuously release a function of a dataset that is
revealed one element at a time. The challenge is to maintain a good
approximation while keeping the combined output over all time steps
differentially private. In the special case of $\textit{continual counting}$ we
seek to approximate a sum of binary input elements. This problem has received
considerable attention lately, in part due to its relevance in implementations
of differentially private stochastic gradient descent. $\textit{Factorization
mechanisms}$ are the leading approach to continual counting, but the best such
mechanisms do not work well in $\textit{streaming}$ settings since they require
space proportional to the size of the input. In this paper, we present a simple
approach to approximating factorization mechanisms in low space via
$\textit{binning}$, where adjacent matrix entries with similar values are
changed to be identical in such a way that a matrix-vector product can be
maintained in sublinear space. Our approach has provable sublinear space
guarantees for a class of lower triangular matrices whose entries are
monotonically decreasing away from the diagonal. We show empirically that even
with very low space usage we are able to closely match, and sometimes surpass,
the performance of asymptotically optimal factorization mechanisms. Recently,
and independently of our work, Dvijotham et al. have also suggested an approach
to implementing factorization mechanisms in a streaming setting. Their work
differs from ours in several respects: It only addresses factorization into
$\textit{Toeplitz}$ matrices, only considers $\textit{maximum}$ error, and uses
a different technique based on rational function approximation that seems less
versatile than our binning approach.
| no_new_dataset | 0.939137 |
2412.15527 | Weizhi Xian | Weizhi Xian, Mingliang Zhou, Leong Hou U, Lang Shujun, Bin Fang, Tao
Xiang, Zhaowei Shang, Weijia Jia | PIGUIQA: A Physical Imaging Guided Perceptual Framework for Underwater
Image Quality Assessment | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a Physical Imaging Guided perceptual framework for
Underwater Image Quality Assessment (UIQA), termed PIGUIQA. First, we formulate
UIQA as a comprehensive problem that considers the combined effects of direct
transmission attenuation and backward scattering on image perception. By
leveraging underwater radiative transfer theory, we systematically integrate
physics-based imaging estimations to establish quantitative metrics for these
distortions. Second, recognizing spatial variations in image content
significance and human perceptual sensitivity to distortions, we design a
module built upon a neighborhood attention mechanism for local perception of
images. This module effectively captures subtle features in images, thereby
enhancing the adaptive perception of distortions on the basis of local
information. Third, by employing a global perceptual aggregator that further
integrates holistic image scene with underwater distortion information, the
proposed model accurately predicts image quality scores. Extensive experiments
across multiple benchmarks demonstrate that PIGUIQA achieves state-of-the-art
performance while maintaining robust cross-dataset generalizability. The
implementation is publicly available at
https://anonymous.4open.science/r/PIGUIQA-A465/
| [
{
"version": "v1",
"created": "Fri, 20 Dec 2024 03:31:45 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 03:19:13 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Xian",
"Weizhi",
""
],
[
"Zhou",
"Mingliang",
""
],
[
"U",
"Leong Hou",
""
],
[
"Shujun",
"Lang",
""
],
[
"Fang",
"Bin",
""
],
[
"Xiang",
"Tao",
""
],
[
"Shang",
"Zhaowei",
""
],
[
"Jia",
"Weijia",
""
]
]
| TITLE: PIGUIQA: A Physical Imaging Guided Perceptual Framework for Underwater
Image Quality Assessment
ABSTRACT: In this paper, we propose a Physical Imaging Guided perceptual framework for
Underwater Image Quality Assessment (UIQA), termed PIGUIQA. First, we formulate
UIQA as a comprehensive problem that considers the combined effects of direct
transmission attenuation and backward scattering on image perception. By
leveraging underwater radiative transfer theory, we systematically integrate
physics-based imaging estimations to establish quantitative metrics for these
distortions. Second, recognizing spatial variations in image content
significance and human perceptual sensitivity to distortions, we design a
module built upon a neighborhood attention mechanism for local perception of
images. This module effectively captures subtle features in images, thereby
enhancing the adaptive perception of distortions on the basis of local
information. Third, by employing a global perceptual aggregator that further
integrates holistic image scene with underwater distortion information, the
proposed model accurately predicts image quality scores. Extensive experiments
across multiple benchmarks demonstrate that PIGUIQA achieves state-of-the-art
performance while maintaining robust cross-dataset generalizability. The
implementation is publicly available at
https://anonymous.4open.science/r/PIGUIQA-A465/
| no_new_dataset | 0.947866 |
2412.16490 | Jiayi Chen | Jiayi Chen, Yubin Ke, He Wang | BODex: Scalable and Efficient Robotic Dexterous Grasp Synthesis Using
Bilevel Optimization | ICRA 2025 | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | Robotic dexterous grasping is important for interacting with the environment.
To unleash the potential of data-driven models for dexterous grasping, a
large-scale, high-quality dataset is essential. While gradient-based
optimization offers a promising way for constructing such datasets, previous
works suffer from limitations, such as inefficiency, strong assumptions in the
grasp quality energy, or limited object sets for experiments. Moreover, the
lack of a standard benchmark for comparing different methods and datasets
hinders progress in this field. To address these challenges, we develop a
highly efficient synthesis system and a comprehensive benchmark with MuJoCo for
dexterous grasping. We formulate grasp synthesis as a bilevel optimization
problem, combining a novel lower-level quadratic programming (QP) with an
upper-level gradient descent process. By leveraging recent advances in
CUDA-accelerated robotic libraries and GPU-based QP solvers, our system can
parallelize thousands of grasps and synthesize over 49 grasps per second on a
single 3090 GPU. Our synthesized grasps for Shadow, Allegro, and Leap hands all
achieve a success rate above 75% in simulation, with a penetration depth under
1 mm, outperforming existing baselines on nearly all metrics. Compared to the
previous large-scale dataset, DexGraspNet, our dataset significantly improves
the performance of learning models, with a success rate from around 40% to 80%
in simulation. Real-world testing of the trained model on the Shadow Hand
achieves an 81% success rate across 20 diverse objects. The codes and datasets
are released on our project page: https://pku-epic.github.io/BODex.
| [
{
"version": "v1",
"created": "Sat, 21 Dec 2024 05:22:53 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 03:12:16 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Chen",
"Jiayi",
""
],
[
"Ke",
"Yubin",
""
],
[
"Wang",
"He",
""
]
]
| TITLE: BODex: Scalable and Efficient Robotic Dexterous Grasp Synthesis Using
Bilevel Optimization
ABSTRACT: Robotic dexterous grasping is important for interacting with the environment.
To unleash the potential of data-driven models for dexterous grasping, a
large-scale, high-quality dataset is essential. While gradient-based
optimization offers a promising way for constructing such datasets, previous
works suffer from limitations, such as inefficiency, strong assumptions in the
grasp quality energy, or limited object sets for experiments. Moreover, the
lack of a standard benchmark for comparing different methods and datasets
hinders progress in this field. To address these challenges, we develop a
highly efficient synthesis system and a comprehensive benchmark with MuJoCo for
dexterous grasping. We formulate grasp synthesis as a bilevel optimization
problem, combining a novel lower-level quadratic programming (QP) with an
upper-level gradient descent process. By leveraging recent advances in
CUDA-accelerated robotic libraries and GPU-based QP solvers, our system can
parallelize thousands of grasps and synthesize over 49 grasps per second on a
single 3090 GPU. Our synthesized grasps for Shadow, Allegro, and Leap hands all
achieve a success rate above 75% in simulation, with a penetration depth under
1 mm, outperforming existing baselines on nearly all metrics. Compared to the
previous large-scale dataset, DexGraspNet, our dataset significantly improves
the performance of learning models, with a success rate from around 40% to 80%
in simulation. Real-world testing of the trained model on the Shadow Hand
achieves an 81% success rate across 20 diverse objects. The codes and datasets
are released on our project page: https://pku-epic.github.io/BODex.
| no_new_dataset | 0.915015 |
2412.16880 | Shenghai Yuan | Shenghai Yuan, Boyang Lou, Thien-Minh Nguyen, Pengyu Yin, Muqing Cao,
Xinghang Xu, Jianping Li, Jie Xu, Siyu Chen, Lihua Xie | Large-Scale UWB Anchor Calibration and One-Shot Localization Using
Gaussian Process | This work has been accepted to IEEE International Conference on
Robotics and Automation (ICRA) @ 2025 IEEE. Personal use of this material is
permitted. Permission from IEEE must be obtained for all other uses,
including reprinting/redistribution, creating new works, or reuse of any
copyrighted components of this work in other media | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | Ultra-wideband (UWB) is gaining popularity with devices like AirTags for
precise home item localization but faces significant challenges when scaled to
large environments like seaports. The main challenges are calibration and
localization in obstructed conditions, which are common in logistics
environments. Traditional calibration methods, dependent on line-of-sight
(LoS), are slow, costly, and unreliable in seaports and warehouses, making
large-scale localization a significant pain point in the industry. To overcome
these challenges, we propose a UWB-LiDAR fusion-based calibration and one-shot
localization framework. Our method uses Gaussian Processes to estimate anchor
position from continuous-time LiDAR Inertial Odometry with sampled UWB ranges.
This approach ensures accurate and reliable calibration with just one round of
sampling in large-scale areas, I.e., 600x450 square meter. With the LoS issues,
UWB-only localization can be problematic, even when anchor positions are known.
We demonstrate that by applying a UWB-range filter, the search range for LiDAR
loop closure descriptors is significantly reduced, improving both accuracy and
speed. This concept can be applied to other loop closure detection methods,
enabling cost-effective localization in large-scale warehouses and seaports. It
significantly improves precision in challenging environments where UWB-only and
LiDAR-Inertial methods fall short, as shown in the video
(https://youtu.be/oY8jQKdM7lU). We will open-source our datasets and
calibration codes for community use.
| [
{
"version": "v1",
"created": "Sun, 22 Dec 2024 06:20:59 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 07:11:20 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Yuan",
"Shenghai",
""
],
[
"Lou",
"Boyang",
""
],
[
"Nguyen",
"Thien-Minh",
""
],
[
"Yin",
"Pengyu",
""
],
[
"Cao",
"Muqing",
""
],
[
"Xu",
"Xinghang",
""
],
[
"Li",
"Jianping",
""
],
[
"Xu",
"Jie",
""
],
[
"Chen",
"Siyu",
""
],
[
"Xie",
"Lihua",
""
]
]
| TITLE: Large-Scale UWB Anchor Calibration and One-Shot Localization Using
Gaussian Process
ABSTRACT: Ultra-wideband (UWB) is gaining popularity with devices like AirTags for
precise home item localization but faces significant challenges when scaled to
large environments like seaports. The main challenges are calibration and
localization in obstructed conditions, which are common in logistics
environments. Traditional calibration methods, dependent on line-of-sight
(LoS), are slow, costly, and unreliable in seaports and warehouses, making
large-scale localization a significant pain point in the industry. To overcome
these challenges, we propose a UWB-LiDAR fusion-based calibration and one-shot
localization framework. Our method uses Gaussian Processes to estimate anchor
position from continuous-time LiDAR Inertial Odometry with sampled UWB ranges.
This approach ensures accurate and reliable calibration with just one round of
sampling in large-scale areas, I.e., 600x450 square meter. With the LoS issues,
UWB-only localization can be problematic, even when anchor positions are known.
We demonstrate that by applying a UWB-range filter, the search range for LiDAR
loop closure descriptors is significantly reduced, improving both accuracy and
speed. This concept can be applied to other loop closure detection methods,
enabling cost-effective localization in large-scale warehouses and seaports. It
significantly improves precision in challenging environments where UWB-only and
LiDAR-Inertial methods fall short, as shown in the video
(https://youtu.be/oY8jQKdM7lU). We will open-source our datasets and
calibration codes for community use.
| no_new_dataset | 0.953665 |
2501.04873 | Alexander Gabriel Valverde Guillen | Alexander Valverde and Luis Solano | Back Home: A Machine Learning Approach to Seashell Classification and
Ecosystem Restoration | null | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In Costa Rica, an average of 5 tons of seashells are extracted from
ecosystems annually. Confiscated seashells, cannot be returned to their
ecosystems due to the lack of origin recognition. To address this issue, we
developed a convolutional neural network (CNN) specifically for seashell
identification. We built a dataset from scratch, consisting of approximately
19000 images from the Pacific and Caribbean coasts. Using this dataset, the
model achieved a classification accuracy exceeding 85%. The model has been
integrated into a user-friendly application, which has classified over 36,000
seashells to date, delivering real-time results within 3 seconds per image. To
further enhance the system's accuracy, an anomaly detection mechanism was
incorporated to filter out irrelevant or anomalous inputs, ensuring only valid
seashell images are processed.
| [
{
"version": "v1",
"created": "Wed, 8 Jan 2025 23:07:10 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 17:35:19 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Valverde",
"Alexander",
""
],
[
"Solano",
"Luis",
""
]
]
| TITLE: Back Home: A Machine Learning Approach to Seashell Classification and
Ecosystem Restoration
ABSTRACT: In Costa Rica, an average of 5 tons of seashells are extracted from
ecosystems annually. Confiscated seashells, cannot be returned to their
ecosystems due to the lack of origin recognition. To address this issue, we
developed a convolutional neural network (CNN) specifically for seashell
identification. We built a dataset from scratch, consisting of approximately
19000 images from the Pacific and Caribbean coasts. Using this dataset, the
model achieved a classification accuracy exceeding 85%. The model has been
integrated into a user-friendly application, which has classified over 36,000
seashells to date, delivering real-time results within 3 seconds per image. To
further enhance the system's accuracy, an anomaly detection mechanism was
incorporated to filter out irrelevant or anomalous inputs, ensuring only valid
seashell images are processed.
| new_dataset | 0.96378 |
2501.05880 | Daniel Rossi | Daniel Rossi, Guido Borghi, Roberto Vezzani | TakuNet: an Energy-Efficient CNN for Real-Time Inference on Embedded UAV
systems in Emergency Response Scenarios | 10 pages, 2 figures, 6 tables, Accepted at WACVW 2025, Tucson
(Arizona), United States | In Proceedings of the Winter Conference on Applications of
Computer Vision (WACV), 2025, pp. 376-385 | null | null | cs.CV cs.PF | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Designing efficient neural networks for embedded devices is a critical
challenge, particularly in applications requiring real-time performance, such
as aerial imaging with drones and UAVs for emergency responses. In this work,
we introduce TakuNet, a novel light-weight architecture which employs
techniques such as depth-wise convolutions and an early downsampling stem to
reduce computational complexity while maintaining high accuracy. It leverages
dense connections for fast convergence during training and uses 16-bit
floating-point precision for optimization on embedded hardware accelerators.
Experimental evaluation on two public datasets shows that TakuNet achieves
near-state-of-the-art accuracy in classifying aerial images of emergency
situations, despite its minimal parameter count. Real-world tests on embedded
devices, namely Jetson Orin Nano and Raspberry Pi, confirm TakuNet's
efficiency, achieving more than 650 fps on the 15W Jetson board, making it
suitable for real-time AI processing on resource-constrained platforms and
advancing the applicability of drones in emergency scenarios. The code and
implementation details are publicly released.
| [
{
"version": "v1",
"created": "Fri, 10 Jan 2025 11:32:56 GMT"
},
{
"version": "v2",
"created": "Thu, 16 Jan 2025 20:35:28 GMT"
},
{
"version": "v3",
"created": "Wed, 5 Mar 2025 21:00:34 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Rossi",
"Daniel",
""
],
[
"Borghi",
"Guido",
""
],
[
"Vezzani",
"Roberto",
""
]
]
| TITLE: TakuNet: an Energy-Efficient CNN for Real-Time Inference on Embedded UAV
systems in Emergency Response Scenarios
ABSTRACT: Designing efficient neural networks for embedded devices is a critical
challenge, particularly in applications requiring real-time performance, such
as aerial imaging with drones and UAVs for emergency responses. In this work,
we introduce TakuNet, a novel light-weight architecture which employs
techniques such as depth-wise convolutions and an early downsampling stem to
reduce computational complexity while maintaining high accuracy. It leverages
dense connections for fast convergence during training and uses 16-bit
floating-point precision for optimization on embedded hardware accelerators.
Experimental evaluation on two public datasets shows that TakuNet achieves
near-state-of-the-art accuracy in classifying aerial images of emergency
situations, despite its minimal parameter count. Real-world tests on embedded
devices, namely Jetson Orin Nano and Raspberry Pi, confirm TakuNet's
efficiency, achieving more than 650 fps on the 15W Jetson board, making it
suitable for real-time AI processing on resource-constrained platforms and
advancing the applicability of drones in emergency scenarios. The code and
implementation details are publicly released.
| no_new_dataset | 0.948346 |
2501.13430 | Rui Xu | Rui Xu, Chao Chen, Yue Sun, Parvathinathan Venkitasubramaniam, Sihong
Xie | Wasserstein-regularized Conformal Prediction under General Distribution
Shift | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Conformal prediction yields a prediction set with guaranteed $1-\alpha$
coverage of the true target under the i.i.d. assumption, which may not hold and
lead to a gap between $1-\alpha$ and the actual coverage. Prior studies bound
the gap using total variation distance, which cannot identify the gap changes
under distribution shift at a given $\alpha$. Besides, existing methods are
mostly limited to covariate shift,while general joint distribution shifts are
more common in practice but less researched.In response, we first propose a
Wasserstein distance-based upper bound of the coverage gap and analyze the
bound using probability measure pushforwards between the shifted joint data and
conformal score distributions, enabling a separation of the effect of covariate
and concept shifts over the coverage gap. We exploit the separation to design
an algorithm based on importance weighting and regularized representation
learning (WR-CP) to reduce the Wasserstein bound with a finite-sample error
bound.WR-CP achieves a controllable balance between conformal prediction
accuracy and efficiency. Experiments on six datasets prove that WR-CP can
reduce coverage gaps to $3.2\%$ across different confidence levels and outputs
prediction sets 37$\%$ smaller than the worst-case approach on average.
| [
{
"version": "v1",
"created": "Thu, 23 Jan 2025 07:29:44 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 09:22:38 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Xu",
"Rui",
""
],
[
"Chen",
"Chao",
""
],
[
"Sun",
"Yue",
""
],
[
"Venkitasubramaniam",
"Parvathinathan",
""
],
[
"Xie",
"Sihong",
""
]
]
| TITLE: Wasserstein-regularized Conformal Prediction under General Distribution
Shift
ABSTRACT: Conformal prediction yields a prediction set with guaranteed $1-\alpha$
coverage of the true target under the i.i.d. assumption, which may not hold and
lead to a gap between $1-\alpha$ and the actual coverage. Prior studies bound
the gap using total variation distance, which cannot identify the gap changes
under distribution shift at a given $\alpha$. Besides, existing methods are
mostly limited to covariate shift,while general joint distribution shifts are
more common in practice but less researched.In response, we first propose a
Wasserstein distance-based upper bound of the coverage gap and analyze the
bound using probability measure pushforwards between the shifted joint data and
conformal score distributions, enabling a separation of the effect of covariate
and concept shifts over the coverage gap. We exploit the separation to design
an algorithm based on importance weighting and regularized representation
learning (WR-CP) to reduce the Wasserstein bound with a finite-sample error
bound.WR-CP achieves a controllable balance between conformal prediction
accuracy and efficiency. Experiments on six datasets prove that WR-CP can
reduce coverage gaps to $3.2\%$ across different confidence levels and outputs
prediction sets 37$\%$ smaller than the worst-case approach on average.
| no_new_dataset | 0.946597 |
2501.15361 | Sajjad Ghiasvand | Sajjad Ghiasvand, Mahnoosh Alizadeh, Ramtin Pedarsani | Decentralized Low-Rank Fine-Tuning of Large Language Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | While parameter-efficient fine-tuning (PEFT) techniques like Low-Rank
Adaptation (LoRA) offer computationally efficient adaptations of Large Language
Models (LLMs), their practical deployment often assumes centralized data and
training environments. However, real-world scenarios frequently involve
distributed, privacy-sensitive datasets that require decentralized solutions.
Federated learning (FL) addresses data privacy by coordinating model updates
across clients, but it is typically based on centralized aggregation through a
parameter server, which can introduce bottlenecks and communication
constraints. Decentralized learning, in contrast, eliminates this dependency by
enabling direct collaboration between clients, improving scalability and
efficiency in distributed environments. Despite its advantages, decentralized
LLM fine-tuning remains underexplored. In this work, we propose Dec-LoRA, a
decentralized fine-tuning algorithm for LLMs based on LoRA. Through extensive
experiments on BERT and LLaMA-2 models, we demonstrate that Dec-LoRA achieves
performance comparable to centralized LoRA under various conditions, including
data heterogeneity and quantization constraints. Additionally, we provide a
rigorous theoretical guarantee proving the convergence of our algorithm to a
stationary point for non-convex and smooth loss functions. These findings
highlight the potential of Dec-LoRA for scalable LLM fine-tuning in
decentralized environments.
| [
{
"version": "v1",
"created": "Sun, 26 Jan 2025 01:56:25 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Feb 2025 00:22:42 GMT"
},
{
"version": "v3",
"created": "Wed, 5 Mar 2025 22:09:09 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Ghiasvand",
"Sajjad",
""
],
[
"Alizadeh",
"Mahnoosh",
""
],
[
"Pedarsani",
"Ramtin",
""
]
]
| TITLE: Decentralized Low-Rank Fine-Tuning of Large Language Models
ABSTRACT: While parameter-efficient fine-tuning (PEFT) techniques like Low-Rank
Adaptation (LoRA) offer computationally efficient adaptations of Large Language
Models (LLMs), their practical deployment often assumes centralized data and
training environments. However, real-world scenarios frequently involve
distributed, privacy-sensitive datasets that require decentralized solutions.
Federated learning (FL) addresses data privacy by coordinating model updates
across clients, but it is typically based on centralized aggregation through a
parameter server, which can introduce bottlenecks and communication
constraints. Decentralized learning, in contrast, eliminates this dependency by
enabling direct collaboration between clients, improving scalability and
efficiency in distributed environments. Despite its advantages, decentralized
LLM fine-tuning remains underexplored. In this work, we propose Dec-LoRA, a
decentralized fine-tuning algorithm for LLMs based on LoRA. Through extensive
experiments on BERT and LLaMA-2 models, we demonstrate that Dec-LoRA achieves
performance comparable to centralized LoRA under various conditions, including
data heterogeneity and quantization constraints. Additionally, we provide a
rigorous theoretical guarantee proving the convergence of our algorithm to a
stationary point for non-convex and smooth loss functions. These findings
highlight the potential of Dec-LoRA for scalable LLM fine-tuning in
decentralized environments.
| no_new_dataset | 0.947235 |
2501.17634 | Lucas Lange | Lucas Lange and Ole Borchardt and Erhard Rahm | Federated Learning With Individualized Privacy Through Client Sampling | Accepted at 10th International Conference on Machine Learning
Technologies (ICMLT 2025) | null | null | null | cs.LG cs.CR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With growing concerns about user data collection, individualized privacy has
emerged as a promising solution to balance protection and utility by accounting
for diverse user privacy preferences. Instead of enforcing a uniform level of
anonymization for all users, this approach allows individuals to choose privacy
settings that align with their comfort levels. Building on this idea, we
propose an adapted method for enabling Individualized Differential Privacy
(IDP) in Federated Learning (FL) by handling clients according to their
personal privacy preferences. By extending the SAMPLE algorithm from
centralized settings to FL, we calculate client-specific sampling rates based
on their heterogeneous privacy budgets and integrate them into a modified
IDP-FedAvg algorithm. We test this method under realistic privacy distributions
and multiple datasets. The experimental results demonstrate that our approach
achieves clear improvements over uniform DP baselines, reducing the trade-off
between privacy and utility. Compared to the alternative SCALE method in
related work, which assigns differing noise scales to clients, our method
performs notably better. However, challenges remain for complex tasks with
non-i.i.d. data, primarily stemming from the constraints of the decentralized
setting.
| [
{
"version": "v1",
"created": "Wed, 29 Jan 2025 13:11:21 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 11:17:31 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Lange",
"Lucas",
""
],
[
"Borchardt",
"Ole",
""
],
[
"Rahm",
"Erhard",
""
]
]
| TITLE: Federated Learning With Individualized Privacy Through Client Sampling
ABSTRACT: With growing concerns about user data collection, individualized privacy has
emerged as a promising solution to balance protection and utility by accounting
for diverse user privacy preferences. Instead of enforcing a uniform level of
anonymization for all users, this approach allows individuals to choose privacy
settings that align with their comfort levels. Building on this idea, we
propose an adapted method for enabling Individualized Differential Privacy
(IDP) in Federated Learning (FL) by handling clients according to their
personal privacy preferences. By extending the SAMPLE algorithm from
centralized settings to FL, we calculate client-specific sampling rates based
on their heterogeneous privacy budgets and integrate them into a modified
IDP-FedAvg algorithm. We test this method under realistic privacy distributions
and multiple datasets. The experimental results demonstrate that our approach
achieves clear improvements over uniform DP baselines, reducing the trade-off
between privacy and utility. Compared to the alternative SCALE method in
related work, which assigns differing noise scales to clients, our method
performs notably better. However, challenges remain for complex tasks with
non-i.i.d. data, primarily stemming from the constraints of the decentralized
setting.
| no_new_dataset | 0.947284 |
2502.01262 | EunSol Park | Eun-Sol Park, MiSo Park, Seung Park, Yong-Goo Shin | FSPGD: Rethinking Black-box Attacks on Semantic Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transferability, the ability of adversarial examples crafted for one model to
deceive other models, is crucial for black-box attacks. Despite advancements in
attack methods for semantic segmentation, transferability remains limited,
reducing their effectiveness in real-world applications. To address this, we
introduce the Feature Similarity Projected Gradient Descent (FSPGD) attack, a
novel black-box approach that enhances both attack performance and
transferability. Unlike conventional segmentation attacks that rely on output
predictions for gradient calculation, FSPGD computes gradients from
intermediate layer features. Specifically, our method introduces a loss
function that targets local information by comparing features between clean
images and adversarial examples, while also disrupting contextual information
by accounting for spatial relationships between objects. Experiments on Pascal
VOC 2012 and Cityscapes datasets demonstrate that FSPGD achieves superior
transferability and attack performance, establishing a new state-of-the-art
benchmark. Code is available at https://github.com/KU-AIVS/FSPGD.
| [
{
"version": "v1",
"created": "Mon, 3 Feb 2025 11:36:01 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 14:50:58 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Park",
"Eun-Sol",
""
],
[
"Park",
"MiSo",
""
],
[
"Park",
"Seung",
""
],
[
"Shin",
"Yong-Goo",
""
]
]
| TITLE: FSPGD: Rethinking Black-box Attacks on Semantic Segmentation
ABSTRACT: Transferability, the ability of adversarial examples crafted for one model to
deceive other models, is crucial for black-box attacks. Despite advancements in
attack methods for semantic segmentation, transferability remains limited,
reducing their effectiveness in real-world applications. To address this, we
introduce the Feature Similarity Projected Gradient Descent (FSPGD) attack, a
novel black-box approach that enhances both attack performance and
transferability. Unlike conventional segmentation attacks that rely on output
predictions for gradient calculation, FSPGD computes gradients from
intermediate layer features. Specifically, our method introduces a loss
function that targets local information by comparing features between clean
images and adversarial examples, while also disrupting contextual information
by accounting for spatial relationships between objects. Experiments on Pascal
VOC 2012 and Cityscapes datasets demonstrate that FSPGD achieves superior
transferability and attack performance, establishing a new state-of-the-art
benchmark. Code is available at https://github.com/KU-AIVS/FSPGD.
| no_new_dataset | 0.943971 |
2502.02954 | Ryotaro Kawata | Ryotaro Kawata, Kazusato Oko, Atsushi Nitanda, Taiji Suzuki | Direct Distributional Optimization for Provable Alignment of Diffusion
Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | We introduce a novel alignment method for diffusion models from distribution
optimization perspectives while providing rigorous convergence guarantees. We
first formulate the problem as a generic regularized loss minimization over
probability distributions and directly optimize the distribution using the Dual
Averaging method. Next, we enable sampling from the learned distribution by
approximating its score function via Doob's $h$-transform technique. The
proposed framework is supported by rigorous convergence guarantees and an
end-to-end bound on the sampling error, which imply that when the original
distribution's score is known accurately, the complexity of sampling from
shifted distributions is independent of isoperimetric conditions. This
framework is broadly applicable to general distribution optimization problems,
including alignment tasks in Reinforcement Learning with Human Feedback (RLHF),
Direct Preference Optimization (DPO), and Kahneman-Tversky Optimization (KTO).
We empirically validate its performance on synthetic and image datasets using
the DPO objective.
| [
{
"version": "v1",
"created": "Wed, 5 Feb 2025 07:35:15 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Mar 2025 03:56:38 GMT"
},
{
"version": "v3",
"created": "Thu, 6 Mar 2025 01:25:25 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Kawata",
"Ryotaro",
""
],
[
"Oko",
"Kazusato",
""
],
[
"Nitanda",
"Atsushi",
""
],
[
"Suzuki",
"Taiji",
""
]
]
| TITLE: Direct Distributional Optimization for Provable Alignment of Diffusion
Models
ABSTRACT: We introduce a novel alignment method for diffusion models from distribution
optimization perspectives while providing rigorous convergence guarantees. We
first formulate the problem as a generic regularized loss minimization over
probability distributions and directly optimize the distribution using the Dual
Averaging method. Next, we enable sampling from the learned distribution by
approximating its score function via Doob's $h$-transform technique. The
proposed framework is supported by rigorous convergence guarantees and an
end-to-end bound on the sampling error, which imply that when the original
distribution's score is known accurately, the complexity of sampling from
shifted distributions is independent of isoperimetric conditions. This
framework is broadly applicable to general distribution optimization problems,
including alignment tasks in Reinforcement Learning with Human Feedback (RLHF),
Direct Preference Optimization (DPO), and Kahneman-Tversky Optimization (KTO).
We empirically validate its performance on synthetic and image datasets using
the DPO objective.
| no_new_dataset | 0.9455 |
2502.08177 | Jacob Goldberg | Aaron Fanous and Jacob Goldberg (1), Ank A. Agarwal (1), Joanna Lin
(1), Anson Zhou (1), Roxana Daneshjou (1), Sanmi Koyejo (1) ((1) Stanford
University) | SycEval: Evaluating LLM Sycophancy | 10 pages | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Large language models (LLMs) are increasingly applied in educational,
clinical, and professional settings, but their tendency for sycophancy --
prioritizing user agreement over independent reasoning -- poses risks to
reliability. This study introduces a framework to evaluate sycophantic behavior
in ChatGPT-4o, Claude-Sonnet, and Gemini-1.5-Pro across AMPS (mathematics) and
MedQuad (medical advice) datasets. Sycophantic behavior was observed in 58.19%
of cases, with Gemini exhibiting the highest rate (62.47%) and ChatGPT the
lowest (56.71%). Progressive sycophancy, leading to correct answers, occurred
in 43.52% of cases, while regressive sycophancy, leading to incorrect answers,
was observed in 14.66%. Preemptive rebuttals demonstrated significantly higher
sycophancy rates than in-context rebuttals (61.75% vs. 56.52%, $Z=5.87$,
$p<0.001$), particularly in computational tasks, where regressive sycophancy
increased significantly (preemptive: 8.13%, in-context: 3.54%, $p<0.001$).
Simple rebuttals maximized progressive sycophancy ($Z=6.59$, $p<0.001$), while
citation-based rebuttals exhibited the highest regressive rates ($Z=6.59$,
$p<0.001$). Sycophantic behavior showed high persistence (78.5%, 95% CI:
[77.2%, 79.8%]) regardless of context or model. These findings emphasize the
risks and opportunities of deploying LLMs in structured and dynamic domains,
offering insights into prompt programming and model optimization for safer AI
applications.
| [
{
"version": "v1",
"created": "Wed, 12 Feb 2025 07:32:42 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 00:41:10 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Fanous",
"Aaron",
""
],
[
"Goldberg",
"Jacob",
""
],
[
"Agarwal",
"Ank A.",
""
],
[
"Lin",
"Joanna",
""
],
[
"Zhou",
"Anson",
""
],
[
"Daneshjou",
"Roxana",
""
],
[
"Koyejo",
"Sanmi",
""
]
]
| TITLE: SycEval: Evaluating LLM Sycophancy
ABSTRACT: Large language models (LLMs) are increasingly applied in educational,
clinical, and professional settings, but their tendency for sycophancy --
prioritizing user agreement over independent reasoning -- poses risks to
reliability. This study introduces a framework to evaluate sycophantic behavior
in ChatGPT-4o, Claude-Sonnet, and Gemini-1.5-Pro across AMPS (mathematics) and
MedQuad (medical advice) datasets. Sycophantic behavior was observed in 58.19%
of cases, with Gemini exhibiting the highest rate (62.47%) and ChatGPT the
lowest (56.71%). Progressive sycophancy, leading to correct answers, occurred
in 43.52% of cases, while regressive sycophancy, leading to incorrect answers,
was observed in 14.66%. Preemptive rebuttals demonstrated significantly higher
sycophancy rates than in-context rebuttals (61.75% vs. 56.52%, $Z=5.87$,
$p<0.001$), particularly in computational tasks, where regressive sycophancy
increased significantly (preemptive: 8.13%, in-context: 3.54%, $p<0.001$).
Simple rebuttals maximized progressive sycophancy ($Z=6.59$, $p<0.001$), while
citation-based rebuttals exhibited the highest regressive rates ($Z=6.59$,
$p<0.001$). Sycophantic behavior showed high persistence (78.5%, 95% CI:
[77.2%, 79.8%]) regardless of context or model. These findings emphasize the
risks and opportunities of deploying LLMs in structured and dynamic domains,
offering insights into prompt programming and model optimization for safer AI
applications.
| no_new_dataset | 0.950365 |
2502.12360 | Sujan Sai Gannamaneni | Sujan Sai Gannamaneni, Rohil Prakash Rao, Michael Mock, Maram Akila,
Stefan Wrobel | Detecting Systematic Weaknesses in Vision Models along Predefined
Human-Understandable Dimensions | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Slice discovery methods (SDMs) are prominent algorithms for finding
systematic weaknesses in DNNs. They identify top-k semantically coherent
slices/subsets of data where a DNN-under-test has low performance. For being
directly useful, slices should be aligned with human-understandable and
relevant dimensions, which, for example, are defined by safety and domain
experts as part of the operational design domain (ODD). While SDMs can be
applied effectively on structured data, their application on image data is
complicated by the lack of semantic metadata. To address these issues, we
present an algorithm that combines foundation models for zero-shot image
classification to generate semantic metadata with methods for combinatorial
search to find systematic weaknesses in images. In contrast to existing
approaches, ours identifies weak slices that are in line with pre-defined
human-understandable dimensions. As the algorithm includes foundation models,
its intermediate and final results may not always be exact. Therefore, we
include an approach to address the impact of noisy metadata. We validate our
algorithm on both synthetic and real-world datasets, demonstrating its ability
to recover human-understandable systematic weaknesses. Furthermore, using our
approach, we identify systematic weaknesses of multiple pre-trained and
publicly available state-of-the-art computer vision DNNs.
| [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 22:50:45 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 18:07:00 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Gannamaneni",
"Sujan Sai",
""
],
[
"Rao",
"Rohil Prakash",
""
],
[
"Mock",
"Michael",
""
],
[
"Akila",
"Maram",
""
],
[
"Wrobel",
"Stefan",
""
]
]
| TITLE: Detecting Systematic Weaknesses in Vision Models along Predefined
Human-Understandable Dimensions
ABSTRACT: Slice discovery methods (SDMs) are prominent algorithms for finding
systematic weaknesses in DNNs. They identify top-k semantically coherent
slices/subsets of data where a DNN-under-test has low performance. For being
directly useful, slices should be aligned with human-understandable and
relevant dimensions, which, for example, are defined by safety and domain
experts as part of the operational design domain (ODD). While SDMs can be
applied effectively on structured data, their application on image data is
complicated by the lack of semantic metadata. To address these issues, we
present an algorithm that combines foundation models for zero-shot image
classification to generate semantic metadata with methods for combinatorial
search to find systematic weaknesses in images. In contrast to existing
approaches, ours identifies weak slices that are in line with pre-defined
human-understandable dimensions. As the algorithm includes foundation models,
its intermediate and final results may not always be exact. Therefore, we
include an approach to address the impact of noisy metadata. We validate our
algorithm on both synthetic and real-world datasets, demonstrating its ability
to recover human-understandable systematic weaknesses. Furthermore, using our
approach, we identify systematic weaknesses of multiple pre-trained and
publicly available state-of-the-art computer vision DNNs.
| no_new_dataset | 0.948251 |
2502.13524 | Wei Dai | Wei Dai, Jun Liu | MobileViM: A Light-weight and Dimension-independent Vision Mamba for 3D
Medical Image Analysis | The corresponding author disagrees with the manuscript submitted to
arXiv | null | null | null | cs.CV cs.AI cs.LG cs.NI | http://creativecommons.org/licenses/by-sa/4.0/ | Efficient evaluation of three-dimensional (3D) medical images is crucial for
diagnostic and therapeutic practices in healthcare. Recent years have seen a
substantial uptake in applying deep learning and computer vision to analyse and
interpret medical images. Traditional approaches, such as convolutional neural
networks (CNNs) and vision transformers (ViTs), face significant computational
challenges, prompting the need for architectural advancements. Recent efforts
have led to the introduction of novel architectures like the ``Mamba'' model as
alternative solutions to traditional CNNs or ViTs. The Mamba model excels in
the linear processing of one-dimensional data with low computational demands.
However, Mamba's potential for 3D medical image analysis remains underexplored
and could face significant computational challenges as the dimension increases.
This manuscript presents MobileViM, a streamlined architecture for efficient
segmentation of 3D medical images. In the MobileViM network, we invent a new
dimension-independent mechanism and a dual-direction traversing approach to
incorporate with a vision-Mamba-based framework. MobileViM also features a
cross-scale bridging technique to improve efficiency and accuracy across
various medical imaging modalities. With these enhancements, MobileViM achieves
segmentation speeds exceeding 90 frames per second (FPS) on a single graphics
processing unit (i.e., NVIDIA RTX 4090). This performance is over 24 FPS faster
than the state-of-the-art deep learning models for processing 3D images with
the same computational resources. In addition, experimental evaluations
demonstrate that MobileViM delivers superior performance, with Dice similarity
scores reaching 92.72%, 86.69%, 80.46%, and 77.43% for PENGWIN, BraTS2024,
ATLAS, and Toothfairy2 datasets, respectively, which significantly surpasses
existing models.
| [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 08:21:59 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Mar 2025 14:42:44 GMT"
},
{
"version": "v3",
"created": "Wed, 5 Mar 2025 01:21:38 GMT"
},
{
"version": "v4",
"created": "Thu, 6 Mar 2025 14:27:12 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Dai",
"Wei",
""
],
[
"Liu",
"Jun",
""
]
]
| TITLE: MobileViM: A Light-weight and Dimension-independent Vision Mamba for 3D
Medical Image Analysis
ABSTRACT: Efficient evaluation of three-dimensional (3D) medical images is crucial for
diagnostic and therapeutic practices in healthcare. Recent years have seen a
substantial uptake in applying deep learning and computer vision to analyse and
interpret medical images. Traditional approaches, such as convolutional neural
networks (CNNs) and vision transformers (ViTs), face significant computational
challenges, prompting the need for architectural advancements. Recent efforts
have led to the introduction of novel architectures like the ``Mamba'' model as
alternative solutions to traditional CNNs or ViTs. The Mamba model excels in
the linear processing of one-dimensional data with low computational demands.
However, Mamba's potential for 3D medical image analysis remains underexplored
and could face significant computational challenges as the dimension increases.
This manuscript presents MobileViM, a streamlined architecture for efficient
segmentation of 3D medical images. In the MobileViM network, we invent a new
dimension-independent mechanism and a dual-direction traversing approach to
incorporate with a vision-Mamba-based framework. MobileViM also features a
cross-scale bridging technique to improve efficiency and accuracy across
various medical imaging modalities. With these enhancements, MobileViM achieves
segmentation speeds exceeding 90 frames per second (FPS) on a single graphics
processing unit (i.e., NVIDIA RTX 4090). This performance is over 24 FPS faster
than the state-of-the-art deep learning models for processing 3D images with
the same computational resources. In addition, experimental evaluations
demonstrate that MobileViM delivers superior performance, with Dice similarity
scores reaching 92.72%, 86.69%, 80.46%, and 77.43% for PENGWIN, BraTS2024,
ATLAS, and Toothfairy2 datasets, respectively, which significantly surpasses
existing models.
| no_new_dataset | 0.950686 |
2502.13728 | Mert Cihangiroglu | Marco Arazzi, Mert Cihangiroglu, Serena Nicolazzo, Antonino Nocera | Secure Federated Data Distillation | null | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Dataset Distillation (DD) is a powerful technique for reducing large datasets
into compact, representative synthetic datasets, accelerating Machine Learning
training. However, traditional DD methods operate in a centralized manner,
which poses significant privacy threats and reduces its applicability. To
mitigate these risks, we propose a Secure Federated Data Distillation (SFDD)
framework to decentralize the distillation process while preserving privacy.
Unlike existing Federated Distillation techniques that focus on training global
models with distilled knowledge, our approach aims to produce a distilled
dataset without exposing local contributions. We leverage the
gradient-matching-based distillation method, adapting it for a distributed
setting where clients contribute to the distillation process without sharing
raw data. The central aggregator iteratively refines a synthetic dataset by
integrating client-side updates while ensuring data confidentiality. To make
our approach resilient to inference attacks perpetrated by the server that
could exploit gradient updates to reconstruct private data, we create an
optimized Local Differential Privacy approach, called LDPO-RLD. Furthermore, we
assess the framework's resilience against malicious clients executing backdoor
attacks (such as Doorping) and demonstrate robustness under the assumption of a
sufficient number of participating clients. Our experimental results
demonstrate the effectiveness of SFDD and that the proposed defense concretely
mitigates the identified vulnerabilities, with minimal impact on the
performance of the distilled dataset. By addressing the interplay between
privacy and federation in dataset distillation, this work advances the field of
privacy-preserving Machine Learning making our SFDD framework a viable solution
for sensitive data-sharing applications.
| [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 13:54:44 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 14:07:57 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Arazzi",
"Marco",
""
],
[
"Cihangiroglu",
"Mert",
""
],
[
"Nicolazzo",
"Serena",
""
],
[
"Nocera",
"Antonino",
""
]
]
| TITLE: Secure Federated Data Distillation
ABSTRACT: Dataset Distillation (DD) is a powerful technique for reducing large datasets
into compact, representative synthetic datasets, accelerating Machine Learning
training. However, traditional DD methods operate in a centralized manner,
which poses significant privacy threats and reduces its applicability. To
mitigate these risks, we propose a Secure Federated Data Distillation (SFDD)
framework to decentralize the distillation process while preserving privacy.
Unlike existing Federated Distillation techniques that focus on training global
models with distilled knowledge, our approach aims to produce a distilled
dataset without exposing local contributions. We leverage the
gradient-matching-based distillation method, adapting it for a distributed
setting where clients contribute to the distillation process without sharing
raw data. The central aggregator iteratively refines a synthetic dataset by
integrating client-side updates while ensuring data confidentiality. To make
our approach resilient to inference attacks perpetrated by the server that
could exploit gradient updates to reconstruct private data, we create an
optimized Local Differential Privacy approach, called LDPO-RLD. Furthermore, we
assess the framework's resilience against malicious clients executing backdoor
attacks (such as Doorping) and demonstrate robustness under the assumption of a
sufficient number of participating clients. Our experimental results
demonstrate the effectiveness of SFDD and that the proposed defense concretely
mitigates the identified vulnerabilities, with minimal impact on the
performance of the distilled dataset. By addressing the interplay between
privacy and federation in dataset distillation, this work advances the field of
privacy-preserving Machine Learning making our SFDD framework a viable solution
for sensitive data-sharing applications.
| no_new_dataset | 0.941654 |
2502.14909 | Hieu Xuan Nguyen | Cuong V. Nguyen, Hieu X. Nguyen, Dung D. Pham Minh and Cuong D. Do | Comparing Deep Neural Network for Multi-Label ECG Diagnosis From Scanned
ECG | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Automated ECG diagnosis has seen significant advancements with deep learning
techniques, but real-world applications still face challenges when dealing with
scanned paper ECGs. In this study, we explore multi-label classification of
ECGs extracted from scanned images, moving beyond traditional binary
classification (normal/abnormal). We evaluate the performance of multiple deep
neural network architectures, including AlexNet, VGG, ResNet, and Vision
Transformer, on scanned ECG datasets. Our comparative analysis examines model
accuracy, robustness to image artifacts, and generalizability across different
ECG conditions. Additionally, we investigate whether ECG signals extracted from
scanned images retain sufficient diagnostic information for reliable automated
classification. The findings highlight the strengths and limitations of each
architecture, providing insights into the feasibility of image-based ECG
diagnosis and its potential integration into clinical workflows.
| [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 02:56:27 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 05:18:12 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Nguyen",
"Cuong V.",
""
],
[
"Nguyen",
"Hieu X.",
""
],
[
"Minh",
"Dung D. Pham",
""
],
[
"Do",
"Cuong D.",
""
]
]
| TITLE: Comparing Deep Neural Network for Multi-Label ECG Diagnosis From Scanned
ECG
ABSTRACT: Automated ECG diagnosis has seen significant advancements with deep learning
techniques, but real-world applications still face challenges when dealing with
scanned paper ECGs. In this study, we explore multi-label classification of
ECGs extracted from scanned images, moving beyond traditional binary
classification (normal/abnormal). We evaluate the performance of multiple deep
neural network architectures, including AlexNet, VGG, ResNet, and Vision
Transformer, on scanned ECG datasets. Our comparative analysis examines model
accuracy, robustness to image artifacts, and generalizability across different
ECG conditions. Additionally, we investigate whether ECG signals extracted from
scanned images retain sufficient diagnostic information for reliable automated
classification. The findings highlight the strengths and limitations of each
architecture, providing insights into the feasibility of image-based ECG
diagnosis and its potential integration into clinical workflows.
| no_new_dataset | 0.948442 |
2502.16600 | Guangliang Liu | Guangliang Liu, Lei Jiang, Xitong Zhang, Kristen Marie Johnson | Diagnosing Moral Reasoning Acquisition in Language Models: Pragmatics
and Generalization | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ensuring that Large Language Models (LLMs) return just responses which adhere
to societal values is crucial for their broader application. Prior research has
shown that LLMs often fail to perform satisfactorily on tasks requiring moral
cognizance, such as ethics-based judgments. While current approaches have
focused on fine-tuning LLMs with curated datasets to improve their capabilities
on such tasks, choosing the optimal learning paradigm to enhance the ethical
responses of LLMs remains an open research debate. In this work, we aim to
address this fundamental question: can current learning paradigms enable LLMs
to acquire sufficient moral reasoning capabilities? Drawing from distributional
semantics theory and the pragmatic nature of moral discourse, our analysis
indicates that performance improvements follow a mechanism similar to that of
semantic-level tasks, and therefore remain affected by the pragmatic nature of
morals latent in discourse, a phenomenon we name the pragmatic dilemma. We
conclude that this pragmatic dilemma imposes significant limitations on the
generalization ability of current learning paradigms, making it the primary
bottleneck for moral reasoning acquisition in LLMs.
| [
{
"version": "v1",
"created": "Sun, 23 Feb 2025 15:00:53 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Feb 2025 04:13:25 GMT"
},
{
"version": "v3",
"created": "Tue, 4 Mar 2025 17:23:23 GMT"
},
{
"version": "v4",
"created": "Thu, 6 Mar 2025 17:56:40 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Liu",
"Guangliang",
""
],
[
"Jiang",
"Lei",
""
],
[
"Zhang",
"Xitong",
""
],
[
"Johnson",
"Kristen Marie",
""
]
]
| TITLE: Diagnosing Moral Reasoning Acquisition in Language Models: Pragmatics
and Generalization
ABSTRACT: Ensuring that Large Language Models (LLMs) return just responses which adhere
to societal values is crucial for their broader application. Prior research has
shown that LLMs often fail to perform satisfactorily on tasks requiring moral
cognizance, such as ethics-based judgments. While current approaches have
focused on fine-tuning LLMs with curated datasets to improve their capabilities
on such tasks, choosing the optimal learning paradigm to enhance the ethical
responses of LLMs remains an open research debate. In this work, we aim to
address this fundamental question: can current learning paradigms enable LLMs
to acquire sufficient moral reasoning capabilities? Drawing from distributional
semantics theory and the pragmatic nature of moral discourse, our analysis
indicates that performance improvements follow a mechanism similar to that of
semantic-level tasks, and therefore remain affected by the pragmatic nature of
morals latent in discourse, a phenomenon we name the pragmatic dilemma. We
conclude that this pragmatic dilemma imposes significant limitations on the
generalization ability of current learning paradigms, making it the primary
bottleneck for moral reasoning acquisition in LLMs.
| no_new_dataset | 0.945801 |
2502.17504 | Yijia Xiao | Yijia Xiao, Wanjia Zhao, Junkai Zhang, Yiqiao Jin, Han Zhang, Zhicheng
Ren, Renliang Sun, Haixin Wang, Guancheng Wan, Pan Lu, Xiao Luo, Yu Zhang,
James Zou, Yizhou Sun, Wei Wang | Protein Large Language Models: A Comprehensive Survey | 24 pages, 4 figures, 5 tables | null | null | null | q-bio.BM cs.AI cs.CE cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Protein-specific large language models (Protein LLMs) are revolutionizing
protein science by enabling more efficient protein structure prediction,
function annotation, and design. While existing surveys focus on specific
aspects or applications, this work provides the first comprehensive overview of
Protein LLMs, covering their architectures, training datasets, evaluation
metrics, and diverse applications. Through a systematic analysis of over 100
articles, we propose a structured taxonomy of state-of-the-art Protein LLMs,
analyze how they leverage large-scale protein sequence data for improved
accuracy, and explore their potential in advancing protein engineering and
biomedical research. Additionally, we discuss key challenges and future
directions, positioning Protein LLMs as essential tools for scientific
discovery in protein science. Resources are maintained at
https://github.com/Yijia-Xiao/Protein-LLM-Survey.
| [
{
"version": "v1",
"created": "Fri, 21 Feb 2025 19:22:10 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 16:14:45 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Xiao",
"Yijia",
""
],
[
"Zhao",
"Wanjia",
""
],
[
"Zhang",
"Junkai",
""
],
[
"Jin",
"Yiqiao",
""
],
[
"Zhang",
"Han",
""
],
[
"Ren",
"Zhicheng",
""
],
[
"Sun",
"Renliang",
""
],
[
"Wang",
"Haixin",
""
],
[
"Wan",
"Guancheng",
""
],
[
"Lu",
"Pan",
""
],
[
"Luo",
"Xiao",
""
],
[
"Zhang",
"Yu",
""
],
[
"Zou",
"James",
""
],
[
"Sun",
"Yizhou",
""
],
[
"Wang",
"Wei",
""
]
]
| TITLE: Protein Large Language Models: A Comprehensive Survey
ABSTRACT: Protein-specific large language models (Protein LLMs) are revolutionizing
protein science by enabling more efficient protein structure prediction,
function annotation, and design. While existing surveys focus on specific
aspects or applications, this work provides the first comprehensive overview of
Protein LLMs, covering their architectures, training datasets, evaluation
metrics, and diverse applications. Through a systematic analysis of over 100
articles, we propose a structured taxonomy of state-of-the-art Protein LLMs,
analyze how they leverage large-scale protein sequence data for improved
accuracy, and explore their potential in advancing protein engineering and
biomedical research. Additionally, we discuss key challenges and future
directions, positioning Protein LLMs as essential tools for scientific
discovery in protein science. Resources are maintained at
https://github.com/Yijia-Xiao/Protein-LLM-Survey.
| no_new_dataset | 0.946051 |
2502.18049 | Shirong Xu | Hengzhi He, Shirong Xu, Guang Cheng | Golden Ratio Weighting Prevents Model Collapse | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recent studies identified an intriguing phenomenon in recursive generative
model training known as model collapse, where models trained on data generated
by previous models exhibit severe performance degradation. Addressing this
issue and developing more effective training strategies have become central
challenges in generative model research. In this paper, we investigate this
phenomenon theoretically within a novel framework, where generative models are
iteratively trained on a combination of newly collected real data and synthetic
data from the previous training step. To develop an optimal training strategy
for integrating real and synthetic data, we evaluate the performance of a
weighted training scheme in various scenarios, including Gaussian distribution
estimation and linear regression. We theoretically characterize the impact of
the mixing proportion and weighting scheme of synthetic data on the final
model's performance. Our key finding is that, across different settings, the
optimal weighting scheme under different proportions of synthetic data
asymptotically follows a unified expression, revealing a fundamental trade-off
between leveraging synthetic data and generative model performance. Notably, in
some cases, the optimal weight assigned to real data corresponds to the
reciprocal of the golden ratio. Finally, we validate our theoretical results on
extensive simulated datasets and a real tabular dataset.
| [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 10:15:16 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 16:03:59 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"He",
"Hengzhi",
""
],
[
"Xu",
"Shirong",
""
],
[
"Cheng",
"Guang",
""
]
]
| TITLE: Golden Ratio Weighting Prevents Model Collapse
ABSTRACT: Recent studies identified an intriguing phenomenon in recursive generative
model training known as model collapse, where models trained on data generated
by previous models exhibit severe performance degradation. Addressing this
issue and developing more effective training strategies have become central
challenges in generative model research. In this paper, we investigate this
phenomenon theoretically within a novel framework, where generative models are
iteratively trained on a combination of newly collected real data and synthetic
data from the previous training step. To develop an optimal training strategy
for integrating real and synthetic data, we evaluate the performance of a
weighted training scheme in various scenarios, including Gaussian distribution
estimation and linear regression. We theoretically characterize the impact of
the mixing proportion and weighting scheme of synthetic data on the final
model's performance. Our key finding is that, across different settings, the
optimal weighting scheme under different proportions of synthetic data
asymptotically follows a unified expression, revealing a fundamental trade-off
between leveraging synthetic data and generative model performance. Notably, in
some cases, the optimal weight assigned to real data corresponds to the
reciprocal of the golden ratio. Finally, we validate our theoretical results on
extensive simulated datasets and a real tabular dataset.
| no_new_dataset | 0.949902 |
2502.18232 | Jun Zeng | Jun Zeng, Debesh Jha, Ertugrul Aktas, Elif Keles, Alpay
Medetalibeyoglu, Matthew Antalek, Robert Lewandowski, Daniela Ladner, Amir A.
Borhani, Gorkem Durak, Ulas Bagci | A Reverse Mamba Attention Network for Pathological Liver Segmentation | 8 pages, 3 figures | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present RMA-Mamba, a novel architecture that advances the capabilities of
vision state space models through a specialized reverse mamba attention module
(RMA). The key innovation lies in RMA-Mamba's ability to capture long-range
dependencies while maintaining precise local feature representation through its
hierarchical processing pipeline. By integrating Vision Mamba (VMamba)'s
efficient sequence modeling with RMA's targeted feature refinement, our
architecture achieves superior feature learning across multiple scales. This
dual-mechanism approach enables robust handling of complex morphological
patterns while maintaining computational efficiency. We demonstrate RMA-Mamba's
effectiveness in the challenging domain of pathological liver segmentation
(from both CT and MRI), where traditional segmentation approaches often fail
due to tissue variations. When evaluated on a newly introduced cirrhotic liver
dataset (CirrMRI600+) of T2-weighted MRI scans, RMA-Mamba achieves the
state-of-the-art performance with a Dice coefficient of 92.08%, mean IoU of
87.36%, and recall of 92.96%. The architecture's generalizability is further
validated on the cancerous liver segmentation from CT scans (LiTS: Liver Tumor
Segmentation dataset), yielding a Dice score of 92.9% and mIoU of 88.99%. Our
code is available for public: https://github.com/JunZengz/RMAMamba.
| [
{
"version": "v1",
"created": "Sun, 23 Feb 2025 20:41:25 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 19:18:00 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Zeng",
"Jun",
""
],
[
"Jha",
"Debesh",
""
],
[
"Aktas",
"Ertugrul",
""
],
[
"Keles",
"Elif",
""
],
[
"Medetalibeyoglu",
"Alpay",
""
],
[
"Antalek",
"Matthew",
""
],
[
"Lewandowski",
"Robert",
""
],
[
"Ladner",
"Daniela",
""
],
[
"Borhani",
"Amir A.",
""
],
[
"Durak",
"Gorkem",
""
],
[
"Bagci",
"Ulas",
""
]
]
| TITLE: A Reverse Mamba Attention Network for Pathological Liver Segmentation
ABSTRACT: We present RMA-Mamba, a novel architecture that advances the capabilities of
vision state space models through a specialized reverse mamba attention module
(RMA). The key innovation lies in RMA-Mamba's ability to capture long-range
dependencies while maintaining precise local feature representation through its
hierarchical processing pipeline. By integrating Vision Mamba (VMamba)'s
efficient sequence modeling with RMA's targeted feature refinement, our
architecture achieves superior feature learning across multiple scales. This
dual-mechanism approach enables robust handling of complex morphological
patterns while maintaining computational efficiency. We demonstrate RMA-Mamba's
effectiveness in the challenging domain of pathological liver segmentation
(from both CT and MRI), where traditional segmentation approaches often fail
due to tissue variations. When evaluated on a newly introduced cirrhotic liver
dataset (CirrMRI600+) of T2-weighted MRI scans, RMA-Mamba achieves the
state-of-the-art performance with a Dice coefficient of 92.08%, mean IoU of
87.36%, and recall of 92.96%. The architecture's generalizability is further
validated on the cancerous liver segmentation from CT scans (LiTS: Liver Tumor
Segmentation dataset), yielding a Dice score of 92.9% and mIoU of 88.99%. Our
code is available for public: https://github.com/JunZengz/RMAMamba.
| new_dataset | 0.946448 |
2502.19660 | Zikuan Li | Zikuan Li, Qiaoyun Wu, Jialin Zhang, Kaijun Zhang, Jun Wang | Noise-Injected Spiking Graph Convolution for Energy-Efficient 3D Point
Cloud Denoising | Accepted by AAAI 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spiking neural networks (SNNs), inspired by the spiking computation paradigm
of the biological neural systems, have exhibited superior energy efficiency in
2D classification tasks over traditional artificial neural networks (ANNs).
However, the regression potential of SNNs has not been well explored,
especially in 3D point cloud processing. In this paper, we propose
noise-injected spiking graph convolutional networks to leverage the full
regression potential of SNNs in 3D point cloud denoising. Specifically, we
first emulate the noise-injected neuronal dynamics to build noise-injected
spiking neurons. On this basis, we design noise-injected spiking graph
convolution for promoting disturbance-aware spiking representation learning on
3D points. Starting from the spiking graph convolution, we build two SNN-based
denoising networks. One is a purely spiking graph convolutional network, which
achieves low accuracy loss compared with some ANN-based alternatives, while
resulting in significantly reduced energy consumption on two benchmark
datasets, PU-Net and PC-Net. The other is a hybrid architecture that combines
ANN-based learning with a high performance-efficiency trade-off in just a few
time steps. Our work lights up SNN's potential for 3D point cloud denoising,
injecting new perspectives of exploring the deployment on neuromorphic chips
while paving the way for developing energy-efficient 3D data acquisition
devices.
| [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 01:04:23 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 03:14:41 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Li",
"Zikuan",
""
],
[
"Wu",
"Qiaoyun",
""
],
[
"Zhang",
"Jialin",
""
],
[
"Zhang",
"Kaijun",
""
],
[
"Wang",
"Jun",
""
]
]
| TITLE: Noise-Injected Spiking Graph Convolution for Energy-Efficient 3D Point
Cloud Denoising
ABSTRACT: Spiking neural networks (SNNs), inspired by the spiking computation paradigm
of the biological neural systems, have exhibited superior energy efficiency in
2D classification tasks over traditional artificial neural networks (ANNs).
However, the regression potential of SNNs has not been well explored,
especially in 3D point cloud processing. In this paper, we propose
noise-injected spiking graph convolutional networks to leverage the full
regression potential of SNNs in 3D point cloud denoising. Specifically, we
first emulate the noise-injected neuronal dynamics to build noise-injected
spiking neurons. On this basis, we design noise-injected spiking graph
convolution for promoting disturbance-aware spiking representation learning on
3D points. Starting from the spiking graph convolution, we build two SNN-based
denoising networks. One is a purely spiking graph convolutional network, which
achieves low accuracy loss compared with some ANN-based alternatives, while
resulting in significantly reduced energy consumption on two benchmark
datasets, PU-Net and PC-Net. The other is a hybrid architecture that combines
ANN-based learning with a high performance-efficiency trade-off in just a few
time steps. Our work lights up SNN's potential for 3D point cloud denoising,
injecting new perspectives of exploring the deployment on neuromorphic chips
while paving the way for developing energy-efficient 3D data acquisition
devices.
| no_new_dataset | 0.94801 |
2502.20490 | Yicheng Fu | MohammadHossein Rezaei, Yicheng Fu, Phil Cuvin, Caleb Ziems, Yanzhe
Zhang, Hao Zhu, Diyi Yang | EgoNormia: Benchmarking Physical Social Norm Understanding | null | null | null | null | cs.CV cs.AI cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Human activity is moderated by norms. However, machines are often trained
without explicit supervision on norm understanding and reasoning, especially
when the norms are grounded in a physical and social context. To improve and
evaluate the normative reasoning capability of vision-language models (VLMs),
we present EgoNormia $\|\epsilon\|$, consisting of 1,853 ego-centric videos of
human interactions, each of which has two related questions evaluating both the
prediction and justification of normative actions. The normative actions
encompass seven categories: safety, privacy, proxemics, politeness,
cooperation, coordination/proactivity, and communication/legibility. To compile
this dataset at scale, we propose a novel pipeline leveraging video sampling,
automatic answer generation, filtering, and human validation. Our work
demonstrates that current state-of-the-art vision-language models lack robust
norm understanding, scoring a maximum of 45% on EgoNormia (versus a human bench
of 92%). Our analysis of performance in each dimension highlights the
significant risks of safety, privacy, and the lack of collaboration and
communication capability when applied to real-world agents. We additionally
show that through a retrieval-based generation method, it is possible to use
EgoNormia to enhance normative reasoning in VLMs.
| [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 19:54:16 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 00:59:40 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Rezaei",
"MohammadHossein",
""
],
[
"Fu",
"Yicheng",
""
],
[
"Cuvin",
"Phil",
""
],
[
"Ziems",
"Caleb",
""
],
[
"Zhang",
"Yanzhe",
""
],
[
"Zhu",
"Hao",
""
],
[
"Yang",
"Diyi",
""
]
]
| TITLE: EgoNormia: Benchmarking Physical Social Norm Understanding
ABSTRACT: Human activity is moderated by norms. However, machines are often trained
without explicit supervision on norm understanding and reasoning, especially
when the norms are grounded in a physical and social context. To improve and
evaluate the normative reasoning capability of vision-language models (VLMs),
we present EgoNormia $\|\epsilon\|$, consisting of 1,853 ego-centric videos of
human interactions, each of which has two related questions evaluating both the
prediction and justification of normative actions. The normative actions
encompass seven categories: safety, privacy, proxemics, politeness,
cooperation, coordination/proactivity, and communication/legibility. To compile
this dataset at scale, we propose a novel pipeline leveraging video sampling,
automatic answer generation, filtering, and human validation. Our work
demonstrates that current state-of-the-art vision-language models lack robust
norm understanding, scoring a maximum of 45% on EgoNormia (versus a human bench
of 92%). Our analysis of performance in each dimension highlights the
significant risks of safety, privacy, and the lack of collaboration and
communication capability when applied to real-world agents. We additionally
show that through a retrieval-based generation method, it is possible to use
EgoNormia to enhance normative reasoning in VLMs.
| new_dataset | 0.956309 |
2502.20637 | Chen Yuqian | Yuqian Chen, Leo Zekelman, Yui Lo, Suheyla Cetin-Karayumak, Tengfei
Xue, Yogesh Rathi, Nikos Makris, Fan Zhang, Weidong Cai, Lauren J. O'Donnell | TractCloud-FOV: Deep Learning-based Robust Tractography Parcellation in
Diffusion MRI with Incomplete Field of View | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tractography parcellation classifies streamlines reconstructed from diffusion
MRI into anatomically defined fiber tracts for clinical and research
applications. However, clinical scans often have incomplete fields of view
(FOV) where brain regions are partially imaged, leading to partial or truncated
fiber tracts. To address this challenge, we introduce TractCloud-FOV, a deep
learning framework that robustly parcellates tractography under conditions of
incomplete FOV. We propose a novel training strategy, FOV-Cut Augmentation
(FOV-CA), in which we synthetically cut tractograms to simulate a spectrum of
real-world inferior FOV cutoff scenarios. This data augmentation approach
enriches the training set with realistic truncated streamlines, enabling the
model to achieve superior generalization. We evaluate the proposed
TractCloud-FOV on both synthetically cut tractography and two real-life
datasets with incomplete FOV. TractCloud-FOV significantly outperforms several
state-of-the-art methods on all testing datasets in terms of streamline
classification accuracy, generalization ability, tract anatomical depiction,
and computational efficiency. Overall, TractCloud-FOV achieves efficient and
consistent tractography parcellation in diffusion MRI with incomplete FOV.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 01:36:38 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 04:31:21 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Chen",
"Yuqian",
""
],
[
"Zekelman",
"Leo",
""
],
[
"Lo",
"Yui",
""
],
[
"Cetin-Karayumak",
"Suheyla",
""
],
[
"Xue",
"Tengfei",
""
],
[
"Rathi",
"Yogesh",
""
],
[
"Makris",
"Nikos",
""
],
[
"Zhang",
"Fan",
""
],
[
"Cai",
"Weidong",
""
],
[
"O'Donnell",
"Lauren J.",
""
]
]
| TITLE: TractCloud-FOV: Deep Learning-based Robust Tractography Parcellation in
Diffusion MRI with Incomplete Field of View
ABSTRACT: Tractography parcellation classifies streamlines reconstructed from diffusion
MRI into anatomically defined fiber tracts for clinical and research
applications. However, clinical scans often have incomplete fields of view
(FOV) where brain regions are partially imaged, leading to partial or truncated
fiber tracts. To address this challenge, we introduce TractCloud-FOV, a deep
learning framework that robustly parcellates tractography under conditions of
incomplete FOV. We propose a novel training strategy, FOV-Cut Augmentation
(FOV-CA), in which we synthetically cut tractograms to simulate a spectrum of
real-world inferior FOV cutoff scenarios. This data augmentation approach
enriches the training set with realistic truncated streamlines, enabling the
model to achieve superior generalization. We evaluate the proposed
TractCloud-FOV on both synthetically cut tractography and two real-life
datasets with incomplete FOV. TractCloud-FOV significantly outperforms several
state-of-the-art methods on all testing datasets in terms of streamline
classification accuracy, generalization ability, tract anatomical depiction,
and computational efficiency. Overall, TractCloud-FOV achieves efficient and
consistent tractography parcellation in diffusion MRI with incomplete FOV.
| no_new_dataset | 0.953751 |
2503.00168 | Benedikt Blumenstiel | Benedikt Blumenstiel, Nassim Ait Ali Braham, Conrad M Albrecht,
Stefano Maurogiovanni, Paolo Fraccaro | SSL4EO-S12 v1.1: A Multimodal, Multiseasonal Dataset for Pretraining,
Updated | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | This technical report presents SSL4EO-S12 v1.1, a multimodal, multitemporal
Earth Observation dataset designed for pretraining large-scale foundation
models. Building on the success of SSL4EO-S12 v1.0, the new version addresses
the previous challenges of data misalignment and a limited data structure for
low-barrier, analysis-ready EO processing. SSL4EO-S12 v1.1 covers the world's
10,000 largest cities and its surroundings within a 50 km radius across four
seasons, resulting in a diverse collection of nearly one million patches.
SSL4EO-S12 v1.1 packages the data in Zarr file format for cloud-efficient
loading and representation of meta-information such as including cloud masks
and geolocation. Released under the CC-BY-4.0 license, SSL4EO-S12 v1.1
facilitates open research and provides a robust foundation for future
advancements in self-supervised learning and geospatial analysis. The dataset
is available online through https://datapub.fz-juelich.de/ssl4eo-s12, and we
provided additional resources at https://github.com/DLR-MF-DAS/SSL4EO-S12-v1.1.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 20:30:56 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 09:23:35 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Blumenstiel",
"Benedikt",
""
],
[
"Braham",
"Nassim Ait Ali",
""
],
[
"Albrecht",
"Conrad M",
""
],
[
"Maurogiovanni",
"Stefano",
""
],
[
"Fraccaro",
"Paolo",
""
]
]
| TITLE: SSL4EO-S12 v1.1: A Multimodal, Multiseasonal Dataset for Pretraining,
Updated
ABSTRACT: This technical report presents SSL4EO-S12 v1.1, a multimodal, multitemporal
Earth Observation dataset designed for pretraining large-scale foundation
models. Building on the success of SSL4EO-S12 v1.0, the new version addresses
the previous challenges of data misalignment and a limited data structure for
low-barrier, analysis-ready EO processing. SSL4EO-S12 v1.1 covers the world's
10,000 largest cities and its surroundings within a 50 km radius across four
seasons, resulting in a diverse collection of nearly one million patches.
SSL4EO-S12 v1.1 packages the data in Zarr file format for cloud-efficient
loading and representation of meta-information such as including cloud masks
and geolocation. Released under the CC-BY-4.0 license, SSL4EO-S12 v1.1
facilitates open research and provides a robust foundation for future
advancements in self-supervised learning and geospatial analysis. The dataset
is available online through https://datapub.fz-juelich.de/ssl4eo-s12, and we
provided additional resources at https://github.com/DLR-MF-DAS/SSL4EO-S12-v1.1.
| new_dataset | 0.964018 |
2503.00675 | Wenke E | Wenke E, Chao Yuan, Li Li, Yixin Sun, Yona Falinie A. Gaus, Amir
Atapour-Abarghouei, Toby P. Breckon | Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark
for Bird-Eye View Mapping in Autonomous Driving | null | null | null | null | cs.CV cs.RO | http://creativecommons.org/licenses/by/4.0/ | We present Dur360BEV, a novel spherical camera autonomous driving dataset
equipped with a high-resolution 128-channel 3D LiDAR and a RTK-refined GNSS/INS
system, along with a benchmark architecture designed to generate Bird-Eye-View
(BEV) maps using only a single spherical camera. This dataset and benchmark
address the challenges of BEV generation in autonomous driving, particularly by
reducing hardware complexity through the use of a single 360-degree camera
instead of multiple perspective cameras. Within our benchmark architecture, we
propose a novel spherical-image-to-BEV module that leverages spherical imagery
and a refined sampling strategy to project features from 2D to 3D. Our approach
also includes an innovative application of focal loss, specifically adapted to
address the extreme class imbalance often encountered in BEV segmentation
tasks, that demonstrates improved segmentation performance on the Dur360BEV
dataset. The results show that our benchmark not only simplifies the sensor
setup but also achieves competitive performance.
| [
{
"version": "v1",
"created": "Sun, 2 Mar 2025 00:40:50 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 05:59:08 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"E",
"Wenke",
""
],
[
"Yuan",
"Chao",
""
],
[
"Li",
"Li",
""
],
[
"Sun",
"Yixin",
""
],
[
"Gaus",
"Yona Falinie A.",
""
],
[
"Atapour-Abarghouei",
"Amir",
""
],
[
"Breckon",
"Toby P.",
""
]
]
| TITLE: Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark
for Bird-Eye View Mapping in Autonomous Driving
ABSTRACT: We present Dur360BEV, a novel spherical camera autonomous driving dataset
equipped with a high-resolution 128-channel 3D LiDAR and a RTK-refined GNSS/INS
system, along with a benchmark architecture designed to generate Bird-Eye-View
(BEV) maps using only a single spherical camera. This dataset and benchmark
address the challenges of BEV generation in autonomous driving, particularly by
reducing hardware complexity through the use of a single 360-degree camera
instead of multiple perspective cameras. Within our benchmark architecture, we
propose a novel spherical-image-to-BEV module that leverages spherical imagery
and a refined sampling strategy to project features from 2D to 3D. Our approach
also includes an innovative application of focal loss, specifically adapted to
address the extreme class imbalance often encountered in BEV segmentation
tasks, that demonstrates improved segmentation performance on the Dur360BEV
dataset. The results show that our benchmark not only simplifies the sensor
setup but also achieves competitive performance.
| new_dataset | 0.956309 |
2503.00736 | WenHui Lei | Wenhui Lei, Anqi Li, Yusheng Tan, Hanyu Chen, Xiaofan Zhang | Shazam: Unifying Multiple Foundation Models for Advanced Computational
Pathology | 9 pages, 2 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Foundation Models (FMs) in computational pathology (CPath) have significantly
advanced the extraction of meaningful features from histopathology image
datasets, achieving strong performance across various clinical tasks. Despite
their impressive performance, these models often exhibit variability when
applied to different tasks, prompting the need for a unified framework capable
of consistently excelling across various applications. In this work, we propose
Shazam, a novel framework designed to efficiently combine multiple CPath
models. Unlike previous approaches that train a fixed-parameter FM, Shazam
dynamically extracts and refines information from diverse FMs for each specific
task. To ensure that each FM contributes effectively without dominance, a novel
distillation strategy is applied, guiding the student model with features from
all teacher models, which enhances its generalization ability. Experimental
results on two pathology patch classification datasets demonstrate that Shazam
outperforms existing CPath models and other fusion methods. Its lightweight,
flexible design makes it a promising solution for improving CPath analysis in
real-world settings. Code will be available at
https://github.com/Tuner12/Shazam.
| [
{
"version": "v1",
"created": "Sun, 2 Mar 2025 05:20:41 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 03:35:09 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Lei",
"Wenhui",
""
],
[
"Li",
"Anqi",
""
],
[
"Tan",
"Yusheng",
""
],
[
"Chen",
"Hanyu",
""
],
[
"Zhang",
"Xiaofan",
""
]
]
| TITLE: Shazam: Unifying Multiple Foundation Models for Advanced Computational
Pathology
ABSTRACT: Foundation Models (FMs) in computational pathology (CPath) have significantly
advanced the extraction of meaningful features from histopathology image
datasets, achieving strong performance across various clinical tasks. Despite
their impressive performance, these models often exhibit variability when
applied to different tasks, prompting the need for a unified framework capable
of consistently excelling across various applications. In this work, we propose
Shazam, a novel framework designed to efficiently combine multiple CPath
models. Unlike previous approaches that train a fixed-parameter FM, Shazam
dynamically extracts and refines information from diverse FMs for each specific
task. To ensure that each FM contributes effectively without dominance, a novel
distillation strategy is applied, guiding the student model with features from
all teacher models, which enhances its generalization ability. Experimental
results on two pathology patch classification datasets demonstrate that Shazam
outperforms existing CPath models and other fusion methods. Its lightweight,
flexible design makes it a promising solution for improving CPath analysis in
real-world settings. Code will be available at
https://github.com/Tuner12/Shazam.
| no_new_dataset | 0.947137 |
2503.01234 | Sijin Sun | Sijin Sun, Ming Deng, Xingrui Yu, Xinyu Xi, Liangbin Zhao | Self-Adaptive Gamma Context-Aware SSM-based Model for Metal Defect
Detection | 19 pages, 9 figures, under review | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Metal defect detection is critical in industrial quality assurance, yet
existing methods struggle with grayscale variations and complex defect states,
limiting its robustness. To address these challenges, this paper proposes a
Self-Adaptive Gamma Context-Aware SSM-based model(GCM-DET). This advanced
detection framework integrating a Dynamic Gamma Correction (GC) module to
enhance grayscale representation and optimize feature extraction for precise
defect reconstruction. A State-Space Search Management (SSM) architecture
captures robust multi-scale features, effectively handling defects of varying
shapes and scales. Focal Loss is employed to mitigate class imbalance and
refine detection accuracy. Additionally, the CD5-DET dataset is introduced,
specifically designed for port container maintenance, featuring significant
grayscale variations and intricate defect patterns. Experimental results
demonstrate that the proposed model achieves substantial improvements, with
[email protected] gains of 27.6\%, 6.6\%, and 2.6\% on the CD5-DET, NEU-DET, and GC10-DET
datasets.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 06:57:54 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 07:11:32 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Sun",
"Sijin",
""
],
[
"Deng",
"Ming",
""
],
[
"Yu",
"Xingrui",
""
],
[
"Xi",
"Xinyu",
""
],
[
"Zhao",
"Liangbin",
""
]
]
| TITLE: Self-Adaptive Gamma Context-Aware SSM-based Model for Metal Defect
Detection
ABSTRACT: Metal defect detection is critical in industrial quality assurance, yet
existing methods struggle with grayscale variations and complex defect states,
limiting its robustness. To address these challenges, this paper proposes a
Self-Adaptive Gamma Context-Aware SSM-based model(GCM-DET). This advanced
detection framework integrating a Dynamic Gamma Correction (GC) module to
enhance grayscale representation and optimize feature extraction for precise
defect reconstruction. A State-Space Search Management (SSM) architecture
captures robust multi-scale features, effectively handling defects of varying
shapes and scales. Focal Loss is employed to mitigate class imbalance and
refine detection accuracy. Additionally, the CD5-DET dataset is introduced,
specifically designed for port container maintenance, featuring significant
grayscale variations and intricate defect patterns. Experimental results
demonstrate that the proposed model achieves substantial improvements, with
[email protected] gains of 27.6\%, 6.6\%, and 2.6\% on the CD5-DET, NEU-DET, and GC10-DET
datasets.
| new_dataset | 0.957873 |
2503.02118 | C\'edric Solenthaler | C\'edric Solenthaler, Joshua Smailes, Martin Strohmeier | OrbID: Identifying Orbcomm Satellite RF Fingerprints | null | null | 10.14722/spacesec.2025.23031 | null | eess.SP cs.CR | http://creativecommons.org/licenses/by/4.0/ | An increase in availability of Software Defined Radios (SDRs) has caused a
dramatic shift in the threat landscape of legacy satellite systems, opening
them up to easy spoofing attacks by low-budget adversaries. Physical-layer
authentication methods can help improve the security of these systems by
providing additional validation without modifying the space segment. This paper
extends previous research on Radio Frequency Fingerprinting (RFF) of satellite
communication to the Orbcomm satellite formation. The GPS and Iridium
constellations are already well covered in prior research, but the feasibility
of transferring techniques to other formations has not yet been examined, and
raises previously undiscussed challenges.
In this paper, we collect a novel dataset containing 8992474 packets from the
Orbcom satellite constellation using different SDRs and locations. We use this
dataset to train RFF systems based on convolutional neural networks. We achieve
an ROC AUC score of 0.53 when distinguishing different satellites within the
constellation, and 0.98 when distinguishing legitimate satellites from SDRs in
a spoofing scenario. We also demonstrate the possibility of mixing datasets
using different SDRs in different physical locations.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 23:00:32 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 14:33:31 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Solenthaler",
"Cédric",
""
],
[
"Smailes",
"Joshua",
""
],
[
"Strohmeier",
"Martin",
""
]
]
| TITLE: OrbID: Identifying Orbcomm Satellite RF Fingerprints
ABSTRACT: An increase in availability of Software Defined Radios (SDRs) has caused a
dramatic shift in the threat landscape of legacy satellite systems, opening
them up to easy spoofing attacks by low-budget adversaries. Physical-layer
authentication methods can help improve the security of these systems by
providing additional validation without modifying the space segment. This paper
extends previous research on Radio Frequency Fingerprinting (RFF) of satellite
communication to the Orbcomm satellite formation. The GPS and Iridium
constellations are already well covered in prior research, but the feasibility
of transferring techniques to other formations has not yet been examined, and
raises previously undiscussed challenges.
In this paper, we collect a novel dataset containing 8992474 packets from the
Orbcom satellite constellation using different SDRs and locations. We use this
dataset to train RFF systems based on convolutional neural networks. We achieve
an ROC AUC score of 0.53 when distinguishing different satellites within the
constellation, and 0.98 when distinguishing legitimate satellites from SDRs in
a spoofing scenario. We also demonstrate the possibility of mixing datasets
using different SDRs in different physical locations.
| new_dataset | 0.96051 |
2503.02356 | Hongtao Xu | Xiulong Yuan, Hongtao Xu, Wenting Shen, Ang Wang, Xiafei Qiu, Jie
Zhang, Yuqiong Liu, Bowen Yu, Junyang Lin, Mingzhen Li, Weile Jia, Yong Li,
Wei Lin | Efficient Long Context Fine-tuning with Chunk Flow | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Long context fine-tuning of large language models(LLMs) involves training on
datasets that are predominantly composed of short sequences and a small
proportion of longer sequences. However, existing approaches overlook this
long-tail distribution and employ training strategies designed specifically for
long sequences. Moreover, these approaches also fail to address the challenges
posed by variable sequence lengths during distributed training, such as load
imbalance in data parallelism and severe pipeline bubbles in pipeline
parallelism. These issues lead to suboptimal training performance and poor GPU
resource utilization. To tackle these problems, we propose a chunk-centric
training method named ChunkFlow. ChunkFlow reorganizes input sequences into
uniformly sized chunks by consolidating short sequences and splitting longer
ones. This approach achieves optimal computational efficiency and balance among
training inputs. Additionally, ChunkFlow incorporates a state-aware chunk
scheduling mechanism to ensure that the peak memory usage during training is
primarily determined by the chunk size rather than the maximum sequence length
in the dataset. Integrating this scheduling mechanism with existing pipeline
scheduling algorithms further enhances the performance of distributed training.
Experimental results demonstrate that, compared with Megatron-LM, ChunkFlow can
be up to 4.53x faster in the long context fine-tuning of LLMs. Furthermore, we
believe that ChunkFlow serves as an effective solution for a broader range of
scenarios, such as long context continual pre-training, where datasets contain
variable-length sequences.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 07:27:41 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 02:44:16 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Yuan",
"Xiulong",
""
],
[
"Xu",
"Hongtao",
""
],
[
"Shen",
"Wenting",
""
],
[
"Wang",
"Ang",
""
],
[
"Qiu",
"Xiafei",
""
],
[
"Zhang",
"Jie",
""
],
[
"Liu",
"Yuqiong",
""
],
[
"Yu",
"Bowen",
""
],
[
"Lin",
"Junyang",
""
],
[
"Li",
"Mingzhen",
""
],
[
"Jia",
"Weile",
""
],
[
"Li",
"Yong",
""
],
[
"Lin",
"Wei",
""
]
]
| TITLE: Efficient Long Context Fine-tuning with Chunk Flow
ABSTRACT: Long context fine-tuning of large language models(LLMs) involves training on
datasets that are predominantly composed of short sequences and a small
proportion of longer sequences. However, existing approaches overlook this
long-tail distribution and employ training strategies designed specifically for
long sequences. Moreover, these approaches also fail to address the challenges
posed by variable sequence lengths during distributed training, such as load
imbalance in data parallelism and severe pipeline bubbles in pipeline
parallelism. These issues lead to suboptimal training performance and poor GPU
resource utilization. To tackle these problems, we propose a chunk-centric
training method named ChunkFlow. ChunkFlow reorganizes input sequences into
uniformly sized chunks by consolidating short sequences and splitting longer
ones. This approach achieves optimal computational efficiency and balance among
training inputs. Additionally, ChunkFlow incorporates a state-aware chunk
scheduling mechanism to ensure that the peak memory usage during training is
primarily determined by the chunk size rather than the maximum sequence length
in the dataset. Integrating this scheduling mechanism with existing pipeline
scheduling algorithms further enhances the performance of distributed training.
Experimental results demonstrate that, compared with Megatron-LM, ChunkFlow can
be up to 4.53x faster in the long context fine-tuning of LLMs. Furthermore, we
believe that ChunkFlow serves as an effective solution for a broader range of
scenarios, such as long context continual pre-training, where datasets contain
variable-length sequences.
| no_new_dataset | 0.950869 |
2503.02365 | Dana Moukheiber | Lama Moukheiber, Mira Moukheiber, Dana Moukheiiber, Jae-Woo Ju,
Hyung-Chul Lee | EchoQA: A Large Collection of Instruction Tuning Data for Echocardiogram
Reports | NeurIPS SafeGenAI 2024 | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | We introduce a novel question-answering (QA) dataset using echocardiogram
reports sourced from the Medical Information Mart for Intensive Care database.
This dataset is specifically designed to enhance QA systems in cardiology,
consisting of 771,244 QA pairs addressing a wide array of cardiac abnormalities
and their severity. We compare large language models (LLMs), including
open-source and biomedical-specific models for zero-shot evaluation, and
closed-source models for zero-shot and three-shot evaluation. Our results show
that fine-tuning LLMs improves performance across various QA metrics,
validating the value of our dataset. Clinicians also qualitatively evaluate the
best-performing model to assess the LLM responses for correctness. Further, we
conduct fine-grained fairness audits to assess the bias-performance trade-off
of LLMs across various social determinants of health. Our objective is to
propel the field forward by establishing a benchmark for LLM AI agents aimed at
supporting clinicians with cardiac differential diagnoses, thereby reducing the
documentation burden that contributes to clinician burnout and enabling
healthcare professionals to focus more on patient care.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 07:45:45 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 03:29:31 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Moukheiber",
"Lama",
""
],
[
"Moukheiber",
"Mira",
""
],
[
"Moukheiiber",
"Dana",
""
],
[
"Ju",
"Jae-Woo",
""
],
[
"Lee",
"Hyung-Chul",
""
]
]
| TITLE: EchoQA: A Large Collection of Instruction Tuning Data for Echocardiogram
Reports
ABSTRACT: We introduce a novel question-answering (QA) dataset using echocardiogram
reports sourced from the Medical Information Mart for Intensive Care database.
This dataset is specifically designed to enhance QA systems in cardiology,
consisting of 771,244 QA pairs addressing a wide array of cardiac abnormalities
and their severity. We compare large language models (LLMs), including
open-source and biomedical-specific models for zero-shot evaluation, and
closed-source models for zero-shot and three-shot evaluation. Our results show
that fine-tuning LLMs improves performance across various QA metrics,
validating the value of our dataset. Clinicians also qualitatively evaluate the
best-performing model to assess the LLM responses for correctness. Further, we
conduct fine-grained fairness audits to assess the bias-performance trade-off
of LLMs across various social determinants of health. Our objective is to
propel the field forward by establishing a benchmark for LLM AI agents aimed at
supporting clinicians with cardiac differential diagnoses, thereby reducing the
documentation burden that contributes to clinician burnout and enabling
healthcare professionals to focus more on patient care.
| new_dataset | 0.959383 |
2503.02448 | Qiyi Wang | Qiyi Wang, Yinning Shao, Yunlong Ma, Min Liu | NodeNAS: Node-Specific Graph Neural Architecture Search for
Out-of-Distribution Generalization | Accepted by DASFAA2025 | null | null | null | cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph neural architecture search (GraphNAS) has demonstrated advantages in
mitigating performance degradation of graph neural networks (GNNs) due to
distribution shifts. Recent approaches introduce weight sharing across tailored
architectures, generating unique GNN architectures for each graph end-to-end.
However, existing GraphNAS methods do not account for distribution patterns
across different graphs and heavily rely on extensive training data. With
sparse or single training graphs, these methods struggle to discover optimal
mappings between graphs and architectures, failing to generalize to
out-of-distribution (OOD) data. In this paper, we propose node-specific graph
neural architecture search(NodeNAS), which aims to tailor distinct aggregation
methods for different nodes through disentangling node topology and graph
distribution with limited datasets. We further propose adaptive aggregation
attention based Multi-dim NodeNAS method(MNNAS), which learns an node-specific
architecture customizer with good generalizability. Specifically, we extend the
vertical depth of the search space, supporting simultaneous node-specific
architecture customization across multiple dimensions. Moreover, we model the
power-law distribution of node degrees under varying assortativity, encoding
structure invariant information to guide architecture customization across each
dimension. Extensive experiments across supervised and unsupervised tasks
demonstrate that MNNAS surpasses state-of-the-art algorithms and achieves
excellent OOD generalization.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 09:45:27 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 02:31:06 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"Qiyi",
""
],
[
"Shao",
"Yinning",
""
],
[
"Ma",
"Yunlong",
""
],
[
"Liu",
"Min",
""
]
]
| TITLE: NodeNAS: Node-Specific Graph Neural Architecture Search for
Out-of-Distribution Generalization
ABSTRACT: Graph neural architecture search (GraphNAS) has demonstrated advantages in
mitigating performance degradation of graph neural networks (GNNs) due to
distribution shifts. Recent approaches introduce weight sharing across tailored
architectures, generating unique GNN architectures for each graph end-to-end.
However, existing GraphNAS methods do not account for distribution patterns
across different graphs and heavily rely on extensive training data. With
sparse or single training graphs, these methods struggle to discover optimal
mappings between graphs and architectures, failing to generalize to
out-of-distribution (OOD) data. In this paper, we propose node-specific graph
neural architecture search(NodeNAS), which aims to tailor distinct aggregation
methods for different nodes through disentangling node topology and graph
distribution with limited datasets. We further propose adaptive aggregation
attention based Multi-dim NodeNAS method(MNNAS), which learns an node-specific
architecture customizer with good generalizability. Specifically, we extend the
vertical depth of the search space, supporting simultaneous node-specific
architecture customization across multiple dimensions. Moreover, we model the
power-law distribution of node degrees under varying assortativity, encoding
structure invariant information to guide architecture customization across each
dimension. Extensive experiments across supervised and unsupervised tasks
demonstrate that MNNAS surpasses state-of-the-art algorithms and achieves
excellent OOD generalization.
| no_new_dataset | 0.951639 |
2503.02660 | Ivan Sipiran | Isaac Aguirre, Ivan Sipiran, Gabriel Monta\~nana | A dataset-free approach for self-supervised learning of 3D reflectional
symmetries | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we explore a self-supervised model that learns to detect the
symmetry of a single object without requiring a dataset-relying solely on the
input object itself. We hypothesize that the symmetry of an object can be
determined by its intrinsic features, eliminating the need for large datasets
during training. Additionally, we design a self-supervised learning strategy
that removes the necessity of ground truth labels. These two key elements make
our approach both effective and efficient, addressing the prohibitive costs
associated with constructing large, labeled datasets for this task. The novelty
of our method lies in computing features for each point on the object based on
the idea that symmetric points should exhibit similar visual appearances. To
achieve this, we leverage features extracted from a foundational image model to
compute a visual descriptor for the points. This approach equips the point
cloud with visual features that facilitate the optimization of our
self-supervised model. Experimental results demonstrate that our method
surpasses the state-of-the-art models trained on large datasets. Furthermore,
our model is more efficient, effective, and operates with minimal computational
and data resources.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 14:22:08 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Mar 2025 19:36:48 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Aguirre",
"Isaac",
""
],
[
"Sipiran",
"Ivan",
""
],
[
"Montañana",
"Gabriel",
""
]
]
| TITLE: A dataset-free approach for self-supervised learning of 3D reflectional
symmetries
ABSTRACT: In this paper, we explore a self-supervised model that learns to detect the
symmetry of a single object without requiring a dataset-relying solely on the
input object itself. We hypothesize that the symmetry of an object can be
determined by its intrinsic features, eliminating the need for large datasets
during training. Additionally, we design a self-supervised learning strategy
that removes the necessity of ground truth labels. These two key elements make
our approach both effective and efficient, addressing the prohibitive costs
associated with constructing large, labeled datasets for this task. The novelty
of our method lies in computing features for each point on the object based on
the idea that symmetric points should exhibit similar visual appearances. To
achieve this, we leverage features extracted from a foundational image model to
compute a visual descriptor for the points. This approach equips the point
cloud with visual features that facilitate the optimization of our
self-supervised model. Experimental results demonstrate that our method
surpasses the state-of-the-art models trained on large datasets. Furthermore,
our model is more efficient, effective, and operates with minimal computational
and data resources.
| no_new_dataset | 0.94545 |
2503.02844 | Paul Janson | Vaibhav Singh, Paul Janson, Paria Mehrbod, Adam Ibrahim, Irina Rish,
Eugene Belilovsky, Benjamin Th\'erien | Beyond Cosine Decay: On the effectiveness of Infinite Learning Rate
Schedule for Continual Pre-training | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The ever-growing availability of unlabeled data presents both opportunities
and challenges for training artificial intelligence systems. While
self-supervised learning (SSL) has emerged as a powerful paradigm for
extracting meaningful representations from vast amounts of unlabeled data,
existing methods still struggle to adapt to the non-stationary, non-IID nature
of real-world data streams without forgetting previously learned knowledge.
Recent works have adopted a repeated cosine annealing schedule for large-scale
continual pre-training; however, these schedules (1) inherently cause
forgetting during the re-warming phase and (2) have not been systematically
compared to existing continual SSL methods. In this work, we systematically
compare the widely used cosine schedule with the recently proposed infinite
learning rate schedule and empirically find the latter to be a more effective
alternative. Our extensive empirical evaluation across diverse image and
language datasets demonstrates that the infinite learning rate schedule
consistently enhances continual pre-training performance compared to a repeated
cosine decay without being restricted to a fixed iteration budget. For
instance, in a small-scale MAE pre-training setup, it outperforms several
strong baselines from the literature. We then scale up our experiments to
larger MAE pre-training and autoregressive language model pre-training. Our
results show that the infinite learning rate schedule remains effective at
scale, surpassing repeated cosine decay for both MAE pre-training and zero-shot
LM benchmarks.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 18:15:57 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 00:17:08 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Singh",
"Vaibhav",
""
],
[
"Janson",
"Paul",
""
],
[
"Mehrbod",
"Paria",
""
],
[
"Ibrahim",
"Adam",
""
],
[
"Rish",
"Irina",
""
],
[
"Belilovsky",
"Eugene",
""
],
[
"Thérien",
"Benjamin",
""
]
]
| TITLE: Beyond Cosine Decay: On the effectiveness of Infinite Learning Rate
Schedule for Continual Pre-training
ABSTRACT: The ever-growing availability of unlabeled data presents both opportunities
and challenges for training artificial intelligence systems. While
self-supervised learning (SSL) has emerged as a powerful paradigm for
extracting meaningful representations from vast amounts of unlabeled data,
existing methods still struggle to adapt to the non-stationary, non-IID nature
of real-world data streams without forgetting previously learned knowledge.
Recent works have adopted a repeated cosine annealing schedule for large-scale
continual pre-training; however, these schedules (1) inherently cause
forgetting during the re-warming phase and (2) have not been systematically
compared to existing continual SSL methods. In this work, we systematically
compare the widely used cosine schedule with the recently proposed infinite
learning rate schedule and empirically find the latter to be a more effective
alternative. Our extensive empirical evaluation across diverse image and
language datasets demonstrates that the infinite learning rate schedule
consistently enhances continual pre-training performance compared to a repeated
cosine decay without being restricted to a fixed iteration budget. For
instance, in a small-scale MAE pre-training setup, it outperforms several
strong baselines from the literature. We then scale up our experiments to
larger MAE pre-training and autoregressive language model pre-training. Our
results show that the infinite learning rate schedule remains effective at
scale, surpassing repeated cosine decay for both MAE pre-training and zero-shot
LM benchmarks.
| no_new_dataset | 0.950365 |
2503.02910 | Wenqi Guo | Wenqi Guo, Yiyang Du, Shan Du | LangGas: Introducing Language in Selective Zero-Shot Background
Subtraction for Semi-Transparent Gas Leak Detection with a New Dataset | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gas leakage poses a significant hazard that requires prevention.
Traditionally, human inspection has been used for detection, a slow and
labour-intensive process. Recent research has applied machine learning
techniques to this problem, yet there remains a shortage of high-quality,
publicly available datasets. This paper introduces a synthetic dataset
featuring diverse backgrounds, interfering foreground objects, diverse leak
locations, and precise segmentation ground truth. We propose a zero-shot method
that combines background subtraction, zero-shot object detection, filtering,
and segmentation to leverage this dataset. Experimental results indicate that
our approach significantly outperforms baseline methods based solely on
background subtraction and zero-shot object detection with segmentation,
reaching an IoU of 69\% overall. We also present an analysis of various prompt
configurations and threshold settings to provide deeper insights into the
performance of our method. The code and dataset will be released after
publication.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 06:17:17 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 05:19:44 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Guo",
"Wenqi",
""
],
[
"Du",
"Yiyang",
""
],
[
"Du",
"Shan",
""
]
]
| TITLE: LangGas: Introducing Language in Selective Zero-Shot Background
Subtraction for Semi-Transparent Gas Leak Detection with a New Dataset
ABSTRACT: Gas leakage poses a significant hazard that requires prevention.
Traditionally, human inspection has been used for detection, a slow and
labour-intensive process. Recent research has applied machine learning
techniques to this problem, yet there remains a shortage of high-quality,
publicly available datasets. This paper introduces a synthetic dataset
featuring diverse backgrounds, interfering foreground objects, diverse leak
locations, and precise segmentation ground truth. We propose a zero-shot method
that combines background subtraction, zero-shot object detection, filtering,
and segmentation to leverage this dataset. Experimental results indicate that
our approach significantly outperforms baseline methods based solely on
background subtraction and zero-shot object detection with segmentation,
reaching an IoU of 69\% overall. We also present an analysis of various prompt
configurations and threshold settings to provide deeper insights into the
performance of our method. The code and dataset will be released after
publication.
| new_dataset | 0.956917 |
2503.03125 | Ziying Song | Ziying Song, Caiyan Jia, Lin Liu, Hongyu Pan, Yongchang Zhang, Junming
Wang, Xingyu Zhang, Shaoqing Xu, Lei Yang, Yadan Luo | Don't Shake the Wheel: Momentum-Aware Planning in End-to-End Autonomous
Driving | 16 pages, 8 figures | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | End-to-end autonomous driving frameworks enable seamless integration of
perception and planning but often rely on one-shot trajectory prediction, which
may lead to unstable control and vulnerability to occlusions in single-frame
perception. To address this, we propose the Momentum-Aware Driving (MomAD)
framework, which introduces trajectory momentum and perception momentum to
stabilize and refine trajectory predictions. MomAD comprises two core
components: (1) Topological Trajectory Matching (TTM) employs Hausdorff
Distance to select the optimal planning query that aligns with prior paths to
ensure coherence;(2) Momentum Planning Interactor (MPI) cross-attends the
selected planning query with historical queries to expand static and dynamic
perception files. This enriched query, in turn, helps regenerate long-horizon
trajectory and reduce collision risks. To mitigate noise arising from dynamic
environments and detection errors, we introduce robust instance denoising
during training, enabling the planning model to focus on critical signals and
improve its robustness. We also propose a novel Trajectory Prediction
Consistency (TPC) metric to quantitatively assess planning stability.
Experiments on the nuScenes dataset demonstrate that MomAD achieves superior
long-term consistency (>=3s) compared to SOTA methods. Moreover, evaluations on
the curated Turning-nuScenes shows that MomAD reduces the collision rate by 26%
and improves TPC by 0.97m (33.45%) over a 6s prediction horizon, while
closedloop on Bench2Drive demonstrates an up to 16.3% improvement in success
rate.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 02:43:52 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 09:53:10 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Song",
"Ziying",
""
],
[
"Jia",
"Caiyan",
""
],
[
"Liu",
"Lin",
""
],
[
"Pan",
"Hongyu",
""
],
[
"Zhang",
"Yongchang",
""
],
[
"Wang",
"Junming",
""
],
[
"Zhang",
"Xingyu",
""
],
[
"Xu",
"Shaoqing",
""
],
[
"Yang",
"Lei",
""
],
[
"Luo",
"Yadan",
""
]
]
| TITLE: Don't Shake the Wheel: Momentum-Aware Planning in End-to-End Autonomous
Driving
ABSTRACT: End-to-end autonomous driving frameworks enable seamless integration of
perception and planning but often rely on one-shot trajectory prediction, which
may lead to unstable control and vulnerability to occlusions in single-frame
perception. To address this, we propose the Momentum-Aware Driving (MomAD)
framework, which introduces trajectory momentum and perception momentum to
stabilize and refine trajectory predictions. MomAD comprises two core
components: (1) Topological Trajectory Matching (TTM) employs Hausdorff
Distance to select the optimal planning query that aligns with prior paths to
ensure coherence;(2) Momentum Planning Interactor (MPI) cross-attends the
selected planning query with historical queries to expand static and dynamic
perception files. This enriched query, in turn, helps regenerate long-horizon
trajectory and reduce collision risks. To mitigate noise arising from dynamic
environments and detection errors, we introduce robust instance denoising
during training, enabling the planning model to focus on critical signals and
improve its robustness. We also propose a novel Trajectory Prediction
Consistency (TPC) metric to quantitatively assess planning stability.
Experiments on the nuScenes dataset demonstrate that MomAD achieves superior
long-term consistency (>=3s) compared to SOTA methods. Moreover, evaluations on
the curated Turning-nuScenes shows that MomAD reduces the collision rate by 26%
and improves TPC by 0.97m (33.45%) over a 6s prediction horizon, while
closedloop on Bench2Drive demonstrates an up to 16.3% improvement in success
rate.
| no_new_dataset | 0.949529 |
2503.03190 | Jingzhou Luo | Jingzhou Luo, Yang Liu, Weixing Chen, Zhen Li, Yaowei Wang, Guanbin
Li, Liang Lin | DSPNet: Dual-vision Scene Perception for Robust 3D Question Answering | Accepted by CVPR 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | 3D Question Answering (3D QA) requires the model to comprehensively
understand its situated 3D scene described by the text, then reason about its
surrounding environment and answer a question under that situation. However,
existing methods usually rely on global scene perception from pure 3D point
clouds and overlook the importance of rich local texture details from
multi-view images. Moreover, due to the inherent noise in camera poses and
complex occlusions, there exists significant feature degradation and reduced
feature robustness problems when aligning 3D point cloud with multi-view
images. In this paper, we propose a Dual-vision Scene Perception Network
(DSPNet), to comprehensively integrate multi-view and point cloud features to
improve robustness in 3D QA. Our Text-guided Multi-view Fusion (TGMF) module
prioritizes image views that closely match the semantic content of the text. To
adaptively fuse back-projected multi-view images with point cloud features, we
design the Adaptive Dual-vision Perception (ADVP) module, enhancing 3D scene
comprehension. Additionally, our Multimodal Context-guided Reasoning (MCGR)
module facilitates robust reasoning by integrating contextual information
across visual and linguistic modalities. Experimental results on SQA3D and
ScanQA datasets demonstrate the superiority of our DSPNet. Codes will be
available at https://github.com/LZ-CH/DSPNet.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 05:13:53 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 03:32:56 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Luo",
"Jingzhou",
""
],
[
"Liu",
"Yang",
""
],
[
"Chen",
"Weixing",
""
],
[
"Li",
"Zhen",
""
],
[
"Wang",
"Yaowei",
""
],
[
"Li",
"Guanbin",
""
],
[
"Lin",
"Liang",
""
]
]
| TITLE: DSPNet: Dual-vision Scene Perception for Robust 3D Question Answering
ABSTRACT: 3D Question Answering (3D QA) requires the model to comprehensively
understand its situated 3D scene described by the text, then reason about its
surrounding environment and answer a question under that situation. However,
existing methods usually rely on global scene perception from pure 3D point
clouds and overlook the importance of rich local texture details from
multi-view images. Moreover, due to the inherent noise in camera poses and
complex occlusions, there exists significant feature degradation and reduced
feature robustness problems when aligning 3D point cloud with multi-view
images. In this paper, we propose a Dual-vision Scene Perception Network
(DSPNet), to comprehensively integrate multi-view and point cloud features to
improve robustness in 3D QA. Our Text-guided Multi-view Fusion (TGMF) module
prioritizes image views that closely match the semantic content of the text. To
adaptively fuse back-projected multi-view images with point cloud features, we
design the Adaptive Dual-vision Perception (ADVP) module, enhancing 3D scene
comprehension. Additionally, our Multimodal Context-guided Reasoning (MCGR)
module facilitates robust reasoning by integrating contextual information
across visual and linguistic modalities. Experimental results on SQA3D and
ScanQA datasets demonstrate the superiority of our DSPNet. Codes will be
available at https://github.com/LZ-CH/DSPNet.
| no_new_dataset | 0.945147 |
2503.03272 | Li Lun | Li Lun, Kunyu Feng, Qinglong Ni, Ling Liang, Yuan Wang, Ying Li,
Dunshan Yu, Xiaoxin Cui | Towards Effective and Sparse Adversarial Attack on Spiking Neural
Networks via Breaking Invisible Surrogate Gradients | Accepted by CVPR 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spiking neural networks (SNNs) have shown their competence in handling
spatial-temporal event-based data with low energy consumption. Similar to
conventional artificial neural networks (ANNs), SNNs are also vulnerable to
gradient-based adversarial attacks, wherein gradients are calculated by
spatial-temporal back-propagation (STBP) and surrogate gradients (SGs).
However, the SGs may be invisible for an inference-only model as they do not
influence the inference results, and current gradient-based attacks are
ineffective for binary dynamic images captured by the dynamic vision sensor
(DVS). While some approaches addressed the issue of invisible SGs through
universal SGs, their SGs lack a correlation with the victim model, resulting in
sub-optimal performance. Moreover, the imperceptibility of existing SNN-based
binary attacks is still insufficient. In this paper, we introduce an innovative
potential-dependent surrogate gradient (PDSG) method to establish a robust
connection between the SG and the model, thereby enhancing the adaptability of
adversarial attacks across various models with invisible SGs. Additionally, we
propose the sparse dynamic attack (SDA) to effectively attack binary dynamic
images. Utilizing a generation-reduction paradigm, SDA can fully optimize the
sparsity of adversarial perturbations. Experimental results demonstrate that
our PDSG and SDA outperform state-of-the-art SNN-based attacks across various
models and datasets. Specifically, our PDSG achieves 100% attack success rate
on ImageNet, and our SDA obtains 82% attack success rate by modifying only
0.24% of the pixels on CIFAR10DVS. The code is available at
https://github.com/ryime/PDSG-SDA .
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 08:52:55 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 13:49:46 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Lun",
"Li",
""
],
[
"Feng",
"Kunyu",
""
],
[
"Ni",
"Qinglong",
""
],
[
"Liang",
"Ling",
""
],
[
"Wang",
"Yuan",
""
],
[
"Li",
"Ying",
""
],
[
"Yu",
"Dunshan",
""
],
[
"Cui",
"Xiaoxin",
""
]
]
| TITLE: Towards Effective and Sparse Adversarial Attack on Spiking Neural
Networks via Breaking Invisible Surrogate Gradients
ABSTRACT: Spiking neural networks (SNNs) have shown their competence in handling
spatial-temporal event-based data with low energy consumption. Similar to
conventional artificial neural networks (ANNs), SNNs are also vulnerable to
gradient-based adversarial attacks, wherein gradients are calculated by
spatial-temporal back-propagation (STBP) and surrogate gradients (SGs).
However, the SGs may be invisible for an inference-only model as they do not
influence the inference results, and current gradient-based attacks are
ineffective for binary dynamic images captured by the dynamic vision sensor
(DVS). While some approaches addressed the issue of invisible SGs through
universal SGs, their SGs lack a correlation with the victim model, resulting in
sub-optimal performance. Moreover, the imperceptibility of existing SNN-based
binary attacks is still insufficient. In this paper, we introduce an innovative
potential-dependent surrogate gradient (PDSG) method to establish a robust
connection between the SG and the model, thereby enhancing the adaptability of
adversarial attacks across various models with invisible SGs. Additionally, we
propose the sparse dynamic attack (SDA) to effectively attack binary dynamic
images. Utilizing a generation-reduction paradigm, SDA can fully optimize the
sparsity of adversarial perturbations. Experimental results demonstrate that
our PDSG and SDA outperform state-of-the-art SNN-based attacks across various
models and datasets. Specifically, our PDSG achieves 100% attack success rate
on ImageNet, and our SDA obtains 82% attack success rate by modifying only
0.24% of the pixels on CIFAR10DVS. The code is available at
https://github.com/ryime/PDSG-SDA .
| no_new_dataset | 0.943138 |
2503.03285 | Thuan Duong | Khoi Anh Nguyen, Linh Yen Vu, Thang Dinh Duong, Thuan Nguyen Duong,
Huy Thanh Nguyen, Vinh Quang Dinh | Enhancing Vietnamese VQA through Curriculum Learning on Raw and
Augmented Text Representations | 10 pages, 3 figures, AAAI-25 Workshop on Document Understanding and
Intelligence | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visual Question Answering (VQA) is a multimodal task requiring reasoning
across textual and visual inputs, which becomes particularly challenging in
low-resource languages like Vietnamese due to linguistic variability and the
lack of high-quality datasets. Traditional methods often rely heavily on
extensive annotated datasets, computationally expensive pipelines, and large
pre-trained models, specifically in the domain of Vietnamese VQA, limiting
their applicability in such scenarios. To address these limitations, we propose
a training framework that combines a paraphrase-based feature augmentation
module with a dynamic curriculum learning strategy. Explicitly, augmented
samples are considered "easy" while raw samples are regarded as "hard". The
framework then utilizes a mechanism that dynamically adjusts the ratio of easy
to hard samples during training, progressively modifying the same dataset to
increase its difficulty level. By enabling gradual adaptation to task
complexity, this approach helps the Vietnamese VQA model generalize well, thus
improving overall performance. Experimental results show consistent
improvements on the OpenViVQA dataset and mixed outcomes on the ViVQA dataset,
highlighting both the potential and challenges of our approach in advancing VQA
for Vietnamese language.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 09:12:16 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 12:42:37 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Nguyen",
"Khoi Anh",
""
],
[
"Vu",
"Linh Yen",
""
],
[
"Duong",
"Thang Dinh",
""
],
[
"Duong",
"Thuan Nguyen",
""
],
[
"Nguyen",
"Huy Thanh",
""
],
[
"Dinh",
"Vinh Quang",
""
]
]
| TITLE: Enhancing Vietnamese VQA through Curriculum Learning on Raw and
Augmented Text Representations
ABSTRACT: Visual Question Answering (VQA) is a multimodal task requiring reasoning
across textual and visual inputs, which becomes particularly challenging in
low-resource languages like Vietnamese due to linguistic variability and the
lack of high-quality datasets. Traditional methods often rely heavily on
extensive annotated datasets, computationally expensive pipelines, and large
pre-trained models, specifically in the domain of Vietnamese VQA, limiting
their applicability in such scenarios. To address these limitations, we propose
a training framework that combines a paraphrase-based feature augmentation
module with a dynamic curriculum learning strategy. Explicitly, augmented
samples are considered "easy" while raw samples are regarded as "hard". The
framework then utilizes a mechanism that dynamically adjusts the ratio of easy
to hard samples during training, progressively modifying the same dataset to
increase its difficulty level. By enabling gradual adaptation to task
complexity, this approach helps the Vietnamese VQA model generalize well, thus
improving overall performance. Experimental results show consistent
improvements on the OpenViVQA dataset and mixed outcomes on the ViVQA dataset,
highlighting both the potential and challenges of our approach in advancing VQA
for Vietnamese language.
| no_new_dataset | 0.946597 |
2503.03370 | Nimra Dilawar | Nimra Dilawar, Sara Nadeem, Javed Iqbal, Waqas Sultani, Mohsen Ali | MIAdapt: Source-free Few-shot Domain Adaptive Object Detection for
Microscopic Images | 6 pages, 5 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Existing generic unsupervised domain adaptation approaches require access to
both a large labeled source dataset and a sufficient unlabeled target dataset
during adaptation. However, collecting a large dataset, even if unlabeled, is a
challenging and expensive endeavor, especially in medical imaging. In addition,
constraints such as privacy issues can result in cases where source data is
unavailable. Taking in consideration these challenges, we propose MIAdapt, an
adaptive approach for Microscopic Imagery Adaptation as a solution for
Source-free Few-shot Domain Adaptive Object detection (SF-FSDA). We also define
two competitive baselines (1) Faster-FreeShot and (2) MT-FreeShot. Extensive
experiments on the challenging M5-Malaria and Raabin-WBC datasets validate the
effectiveness of MIAdapt. Without using any image from the source domain
MIAdapt surpasses state-of-the-art source-free UDA (SF-UDA) methods by +21.3%
mAP and few-shot domain adaptation (FSDA) approaches by +4.7% mAP on
Raabin-WBC. Our code and models will be publicly available.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 10:46:03 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 10:41:28 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Dilawar",
"Nimra",
""
],
[
"Nadeem",
"Sara",
""
],
[
"Iqbal",
"Javed",
""
],
[
"Sultani",
"Waqas",
""
],
[
"Ali",
"Mohsen",
""
]
]
| TITLE: MIAdapt: Source-free Few-shot Domain Adaptive Object Detection for
Microscopic Images
ABSTRACT: Existing generic unsupervised domain adaptation approaches require access to
both a large labeled source dataset and a sufficient unlabeled target dataset
during adaptation. However, collecting a large dataset, even if unlabeled, is a
challenging and expensive endeavor, especially in medical imaging. In addition,
constraints such as privacy issues can result in cases where source data is
unavailable. Taking in consideration these challenges, we propose MIAdapt, an
adaptive approach for Microscopic Imagery Adaptation as a solution for
Source-free Few-shot Domain Adaptive Object detection (SF-FSDA). We also define
two competitive baselines (1) Faster-FreeShot and (2) MT-FreeShot. Extensive
experiments on the challenging M5-Malaria and Raabin-WBC datasets validate the
effectiveness of MIAdapt. Without using any image from the source domain
MIAdapt surpasses state-of-the-art source-free UDA (SF-UDA) methods by +21.3%
mAP and few-shot domain adaptation (FSDA) approaches by +4.7% mAP on
Raabin-WBC. Our code and models will be publicly available.
| no_new_dataset | 0.948822 |
2503.03454 | Chih-Hsun Lin | Ting-Wei Liao, Chih-Hsun Lin, Yu-Lin Tsai, Takao Murakami, Chia-Mu Yu,
Jun Sakuma, Chun-Ying Huang, Hiroaki Kikuchi | Data Poisoning Attacks to Locally Differentially Private Range Query
Protocols | null | null | null | null | cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Local Differential Privacy (LDP) has been widely adopted to protect user
privacy in decentralized data collection. However, recent studies have revealed
that LDP protocols are vulnerable to data poisoning attacks, where malicious
users manipulate their reported data to distort aggregated results. In this
work, we present the first study on data poisoning attacks targeting LDP range
query protocols, focusing on both tree-based and grid-based approaches. We
identify three key challenges in executing such attacks, including crafting
consistent and effective fake data, maintaining data consistency across levels
or grids, and preventing server detection. To address the first two challenges,
we propose novel attack methods that are provably optimal, including a
tree-based attack and a grid-based attack, designed to manipulate range query
results with high effectiveness. \textbf{Our key finding is that the common
post-processing procedure, Norm-Sub, in LDP range query protocols can help the
attacker massively amplify their attack effectiveness.} In addition, we study a
potential countermeasure, but also propose an adaptive attack capable of
evading this defense to address the third challenge. We evaluate our methods
through theoretical analysis and extensive experiments on synthetic and
real-world datasets. Our results show that the proposed attacks can
significantly amplify estimations for arbitrary range queries by manipulating a
small fraction of users, providing 5-10x more influence than a normal user to
the estimation.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 12:40:34 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 14:25:03 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Liao",
"Ting-Wei",
""
],
[
"Lin",
"Chih-Hsun",
""
],
[
"Tsai",
"Yu-Lin",
""
],
[
"Murakami",
"Takao",
""
],
[
"Yu",
"Chia-Mu",
""
],
[
"Sakuma",
"Jun",
""
],
[
"Huang",
"Chun-Ying",
""
],
[
"Kikuchi",
"Hiroaki",
""
]
]
| TITLE: Data Poisoning Attacks to Locally Differentially Private Range Query
Protocols
ABSTRACT: Local Differential Privacy (LDP) has been widely adopted to protect user
privacy in decentralized data collection. However, recent studies have revealed
that LDP protocols are vulnerable to data poisoning attacks, where malicious
users manipulate their reported data to distort aggregated results. In this
work, we present the first study on data poisoning attacks targeting LDP range
query protocols, focusing on both tree-based and grid-based approaches. We
identify three key challenges in executing such attacks, including crafting
consistent and effective fake data, maintaining data consistency across levels
or grids, and preventing server detection. To address the first two challenges,
we propose novel attack methods that are provably optimal, including a
tree-based attack and a grid-based attack, designed to manipulate range query
results with high effectiveness. \textbf{Our key finding is that the common
post-processing procedure, Norm-Sub, in LDP range query protocols can help the
attacker massively amplify their attack effectiveness.} In addition, we study a
potential countermeasure, but also propose an adaptive attack capable of
evading this defense to address the third challenge. We evaluate our methods
through theoretical analysis and extensive experiments on synthetic and
real-world datasets. Our results show that the proposed attacks can
significantly amplify estimations for arbitrary range queries by manipulating a
small fraction of users, providing 5-10x more influence than a normal user to
the estimation.
| no_new_dataset | 0.945751 |
2503.03465 | Fei Zhu | ChenTong Wang, Jincheng Gao, Fei Zhu, Abderrahim Halimi, C\'edric
Richard | DTU-Net: A Multi-Scale Dilated Transformer Network for Nonlinear
Hyperspectral Unmixing | null | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Transformers have shown significant success in hyperspectral unmixing (HU).
However, challenges remain. While multi-scale and long-range spatial
correlations are essential in unmixing tasks, current Transformer-based
unmixing networks, built on Vision Transformer (ViT) or Swin-Transformer,
struggle to capture them effectively. Additionally, current Transformer-based
unmixing networks rely on the linear mixing model, which lacks the flexibility
to accommodate scenarios where nonlinear effects are significant. To address
these limitations, we propose a multi-scale Dilated Transformer-based unmixing
network for nonlinear HU (DTU-Net). The encoder employs two branches. The first
one performs multi-scale spatial feature extraction using Multi-Scale Dilated
Attention (MSDA) in the Dilated Transformer, which varies dilation rates across
attention heads to capture long-range and multi-scale spatial correlations. The
second one performs spectral feature extraction utilizing 3D-CNNs with channel
attention. The outputs from both branches are then fused to integrate
multi-scale spatial and spectral information, which is subsequently transformed
to estimate the abundances. The decoder is designed to accommodate both linear
and nonlinear mixing scenarios. Its interpretability is enhanced by explicitly
modeling the relationships between endmembers, abundances, and nonlinear
coefficients in accordance with the polynomial post-nonlinear mixing model
(PPNMM). Experiments on synthetic and real datasets validate the effectiveness
of the proposed DTU-Net compared to PPNMM-derived methods and several advanced
unmixing networks.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 12:56:33 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 02:55:33 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Wang",
"ChenTong",
""
],
[
"Gao",
"Jincheng",
""
],
[
"Zhu",
"Fei",
""
],
[
"Halimi",
"Abderrahim",
""
],
[
"Richard",
"Cédric",
""
]
]
| TITLE: DTU-Net: A Multi-Scale Dilated Transformer Network for Nonlinear
Hyperspectral Unmixing
ABSTRACT: Transformers have shown significant success in hyperspectral unmixing (HU).
However, challenges remain. While multi-scale and long-range spatial
correlations are essential in unmixing tasks, current Transformer-based
unmixing networks, built on Vision Transformer (ViT) or Swin-Transformer,
struggle to capture them effectively. Additionally, current Transformer-based
unmixing networks rely on the linear mixing model, which lacks the flexibility
to accommodate scenarios where nonlinear effects are significant. To address
these limitations, we propose a multi-scale Dilated Transformer-based unmixing
network for nonlinear HU (DTU-Net). The encoder employs two branches. The first
one performs multi-scale spatial feature extraction using Multi-Scale Dilated
Attention (MSDA) in the Dilated Transformer, which varies dilation rates across
attention heads to capture long-range and multi-scale spatial correlations. The
second one performs spectral feature extraction utilizing 3D-CNNs with channel
attention. The outputs from both branches are then fused to integrate
multi-scale spatial and spectral information, which is subsequently transformed
to estimate the abundances. The decoder is designed to accommodate both linear
and nonlinear mixing scenarios. Its interpretability is enhanced by explicitly
modeling the relationships between endmembers, abundances, and nonlinear
coefficients in accordance with the polynomial post-nonlinear mixing model
(PPNMM). Experiments on synthetic and real datasets validate the effectiveness
of the proposed DTU-Net compared to PPNMM-derived methods and several advanced
unmixing networks.
| no_new_dataset | 0.949389 |
2503.03784 | Kannan Ashwin Viswanathan | Ashwin Viswanathan Kannan, Madhumitha Ganesan | Neural Models of Task Adaptation: A Tutorial on Spiking Networks for
Executive Control | 6 pages | null | null | null | q-bio.NC cs.LG cs.NE | http://creativecommons.org/licenses/by/4.0/ | Understanding cognitive flexibility and task-switching mechanisms in neural
systems requires biologically plausible computational models. This tutorial
presents a step-by-step approach to constructing a spiking neural network (SNN)
that simulates task-switching dynamics within the cognitive control network.
The model incorporates biologically realistic features, including lateral
inhibition, adaptive synaptic weights through unsupervised Spike
Timing-Dependent Plasticity (STDP), and precise neuronal parameterization
within physiologically relevant ranges. The SNN is implemented using Leaky
Integrate-and-Fire (LIF) neurons, which represent excitatory (glutamatergic)
and inhibitory (GABAergic) populations. We utilize two real-world datasets as
tasks, demonstrating how the network learns and dynamically switches between
them. Experimental design follows cognitive psychology paradigms to analyze
neural adaptation, synaptic weight modifications, and emergent behaviors such
as Long-Term Potentiation (LTP), Long-Term Depression (LTD), and Task-Set
Reconfiguration (TSR). Through a series of structured experiments, this
tutorial illustrates how variations in task-switching intervals affect
performance and multitasking efficiency. The results align with empirically
observed neuronal responses, offering insights into the computational
underpinnings of executive function. By following this tutorial, researchers
can develop and extend biologically inspired SNN models for studying cognitive
processes and neural adaptation.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 00:44:34 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Kannan",
"Ashwin Viswanathan",
""
],
[
"Ganesan",
"Madhumitha",
""
]
]
| TITLE: Neural Models of Task Adaptation: A Tutorial on Spiking Networks for
Executive Control
ABSTRACT: Understanding cognitive flexibility and task-switching mechanisms in neural
systems requires biologically plausible computational models. This tutorial
presents a step-by-step approach to constructing a spiking neural network (SNN)
that simulates task-switching dynamics within the cognitive control network.
The model incorporates biologically realistic features, including lateral
inhibition, adaptive synaptic weights through unsupervised Spike
Timing-Dependent Plasticity (STDP), and precise neuronal parameterization
within physiologically relevant ranges. The SNN is implemented using Leaky
Integrate-and-Fire (LIF) neurons, which represent excitatory (glutamatergic)
and inhibitory (GABAergic) populations. We utilize two real-world datasets as
tasks, demonstrating how the network learns and dynamically switches between
them. Experimental design follows cognitive psychology paradigms to analyze
neural adaptation, synaptic weight modifications, and emergent behaviors such
as Long-Term Potentiation (LTP), Long-Term Depression (LTD), and Task-Set
Reconfiguration (TSR). Through a series of structured experiments, this
tutorial illustrates how variations in task-switching intervals affect
performance and multitasking efficiency. The results align with empirically
observed neuronal responses, offering insights into the computational
underpinnings of executive function. By following this tutorial, researchers
can develop and extend biologically inspired SNN models for studying cognitive
processes and neural adaptation.
| no_new_dataset | 0.943971 |
2503.03787 | Gibson Nkhata | Gibson Nkhata Shi Yin Hong and Susan Gauch | Sarcasm Detection as a Catalyst: Improving Stance Detection with
Cross-Target Capabilities | 2 pages, 5 figures, published, published in International Journal On
Advances in Intelligent Systems, volume 17, numbers 3 and 4. arXiv admin
note: text overlap with arXiv:2503.03172 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Stance Detection (SD) has become a critical area of interest due to its
applications in various contexts leading to increased research within NLP. Yet
the subtlety and complexity of texts sourced from online platforms often
containing sarcastic language pose significant challenges for SD algorithms in
accurately determining the authors stance. This paper addresses this by
employing sarcasm for SD. It also tackles the issue of insufficient annotated
data for training SD models on new targets by conducting Cross-Target SD
(CTSD). The proposed approach involves fine-tuning BERT and RoBERTa models
followed by concatenating additional deep learning layers. The approach is
assessed against various State-Of-The-Art baselines for SD demonstrating
superior performance using publicly available datasets. Notably our model
outperforms the best SOTA models on both in-domain SD and CTSD tasks even
before the incorporation of sarcasm-detection pre-training. The integration of
sarcasm knowledge into the model significantly reduces misclassifications of
sarcastic text elements in SD allowing our model to accurately predict 85% of
texts that were previously misclassified without sarcasm-detection pre-training
on in-domain SD. This enhancement contributes to an increase in the models
average macro F1-score. The CTSD task achieves performance comparable to that
of the in-domain task despite using a zero-shot finetuning. We also reveal that
the success of the transfer-learning framework relies on the correlation
between the lexical attributes of sarcasm detection and SD. This study
represents the first exploration of sarcasm detection as an intermediate
transfer-learning task within the context of SD while also leveraging the
concatenation of BERT or RoBERTa with other deep-learning techniques. The
proposed approach establishes a foundational baseline for future research in
this domain.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 05:27:16 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Hong",
"Gibson Nkhata Shi Yin",
""
],
[
"Gauch",
"Susan",
""
]
]
| TITLE: Sarcasm Detection as a Catalyst: Improving Stance Detection with
Cross-Target Capabilities
ABSTRACT: Stance Detection (SD) has become a critical area of interest due to its
applications in various contexts leading to increased research within NLP. Yet
the subtlety and complexity of texts sourced from online platforms often
containing sarcastic language pose significant challenges for SD algorithms in
accurately determining the authors stance. This paper addresses this by
employing sarcasm for SD. It also tackles the issue of insufficient annotated
data for training SD models on new targets by conducting Cross-Target SD
(CTSD). The proposed approach involves fine-tuning BERT and RoBERTa models
followed by concatenating additional deep learning layers. The approach is
assessed against various State-Of-The-Art baselines for SD demonstrating
superior performance using publicly available datasets. Notably our model
outperforms the best SOTA models on both in-domain SD and CTSD tasks even
before the incorporation of sarcasm-detection pre-training. The integration of
sarcasm knowledge into the model significantly reduces misclassifications of
sarcastic text elements in SD allowing our model to accurately predict 85% of
texts that were previously misclassified without sarcasm-detection pre-training
on in-domain SD. This enhancement contributes to an increase in the models
average macro F1-score. The CTSD task achieves performance comparable to that
of the in-domain task despite using a zero-shot finetuning. We also reveal that
the success of the transfer-learning framework relies on the correlation
between the lexical attributes of sarcasm detection and SD. This study
represents the first exploration of sarcasm detection as an intermediate
transfer-learning task within the context of SD while also leveraging the
concatenation of BERT or RoBERTa with other deep-learning techniques. The
proposed approach establishes a foundational baseline for future research in
this domain.
| no_new_dataset | 0.947088 |
2503.03789 | Hiroshi Takahashi | Hiroshi Takahashi, Tomoharu Iwata, Atsutoshi Kumagai, Yuuki Yamanaka,
Tomoya Yamashita | Positive-Unlabeled Diffusion Models for Preventing Sensitive Data
Generation | Accepted at ICLR2025. Code is available at
https://github.com/takahashihiroshi/pudm | null | null | null | cs.LG cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | Diffusion models are powerful generative models but often generate sensitive
data that are unwanted by users, mainly because the unlabeled training data
frequently contain such sensitive data. Since labeling all sensitive data in
the large-scale unlabeled training data is impractical, we address this problem
by using a small amount of labeled sensitive data. In this paper, we propose
positive-unlabeled diffusion models, which prevent the generation of sensitive
data using unlabeled and sensitive data. Our approach can approximate the
evidence lower bound (ELBO) for normal (negative) data using only unlabeled and
sensitive (positive) data. Therefore, even without labeled normal data, we can
maximize the ELBO for normal data and minimize it for labeled sensitive data,
ensuring the generation of only normal data. Through experiments across various
datasets and settings, we demonstrated that our approach can prevent the
generation of sensitive images without compromising image quality.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 07:17:48 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Takahashi",
"Hiroshi",
""
],
[
"Iwata",
"Tomoharu",
""
],
[
"Kumagai",
"Atsutoshi",
""
],
[
"Yamanaka",
"Yuuki",
""
],
[
"Yamashita",
"Tomoya",
""
]
]
| TITLE: Positive-Unlabeled Diffusion Models for Preventing Sensitive Data
Generation
ABSTRACT: Diffusion models are powerful generative models but often generate sensitive
data that are unwanted by users, mainly because the unlabeled training data
frequently contain such sensitive data. Since labeling all sensitive data in
the large-scale unlabeled training data is impractical, we address this problem
by using a small amount of labeled sensitive data. In this paper, we propose
positive-unlabeled diffusion models, which prevent the generation of sensitive
data using unlabeled and sensitive data. Our approach can approximate the
evidence lower bound (ELBO) for normal (negative) data using only unlabeled and
sensitive (positive) data. Therefore, even without labeled normal data, we can
maximize the ELBO for normal data and minimize it for labeled sensitive data,
ensuring the generation of only normal data. Through experiments across various
datasets and settings, we demonstrated that our approach can prevent the
generation of sensitive images without compromising image quality.
| no_new_dataset | 0.953622 |
2503.03794 | Tianyi Huang | Tianyi Huang | Synthetic Data Augmentation for Enhancing Harmful Algal Bloom Detection
with Machine Learning | Accepted Paper at the 2025 IEEE Conference on Technologies for
Sustainability (SusTech) | null | null | null | cs.LG cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Harmful Algal Blooms (HABs) pose severe threats to aquatic ecosystems and
public health, resulting in substantial economic losses globally. Early
detection is crucial but often hindered by the scarcity of high-quality
datasets necessary for training reliable machine learning (ML) models. This
study investigates the use of synthetic data augmentation using Gaussian
Copulas to enhance ML-based HAB detection systems. Synthetic datasets of
varying sizes (100-1,000 samples) were generated using relevant environmental
features$\unicode{x2015}$water temperature, salinity, and UVB
radiation$\unicode{x2015}$with corrected Chlorophyll-a concentration as the
target variable. Experimental results demonstrate that moderate synthetic
augmentation significantly improves model performance (RMSE reduced from 0.4706
to 0.1850; $p < 0.001$). However, excessive synthetic data introduces noise and
reduces predictive accuracy, emphasizing the need for a balanced approach to
data augmentation. These findings highlight the potential of synthetic data to
enhance HAB monitoring systems, offering a scalable and cost-effective method
for early detection and mitigation of ecological and public health risks.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 11:50:04 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Huang",
"Tianyi",
""
]
]
| TITLE: Synthetic Data Augmentation for Enhancing Harmful Algal Bloom Detection
with Machine Learning
ABSTRACT: Harmful Algal Blooms (HABs) pose severe threats to aquatic ecosystems and
public health, resulting in substantial economic losses globally. Early
detection is crucial but often hindered by the scarcity of high-quality
datasets necessary for training reliable machine learning (ML) models. This
study investigates the use of synthetic data augmentation using Gaussian
Copulas to enhance ML-based HAB detection systems. Synthetic datasets of
varying sizes (100-1,000 samples) were generated using relevant environmental
features$\unicode{x2015}$water temperature, salinity, and UVB
radiation$\unicode{x2015}$with corrected Chlorophyll-a concentration as the
target variable. Experimental results demonstrate that moderate synthetic
augmentation significantly improves model performance (RMSE reduced from 0.4706
to 0.1850; $p < 0.001$). However, excessive synthetic data introduces noise and
reduces predictive accuracy, emphasizing the need for a balanced approach to
data augmentation. These findings highlight the potential of synthetic data to
enhance HAB monitoring systems, offering a scalable and cost-effective method
for early detection and mitigation of ecological and public health risks.
| no_new_dataset | 0.950686 |
2503.03797 | Enkhtogtokh Togootogtokh | Enkhtogtokh Togootogtokh, Christian Klasen | VoiceGRPO: Modern MoE Transformers with Group Relative Policy
Optimization GRPO for AI Voice Health Care Applications on Voice Pathology
Detection | null | null | null | null | cs.SD cs.AI eess.AS | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This research introduces a novel AI techniques as Mixture-of-Experts
Transformers with Group Relative Policy Optimization (GRPO) for voice health
care applications on voice pathology detection. With the architectural
innovations, we adopt advanced training paradigms inspired by reinforcement
learning, namely Proximal Policy Optimization (PPO) and Group-wise Regularized
Policy Optimization (GRPO), to enhance model stability and performance.
Experiments conducted on a synthetically generated voice pathology dataset
demonstrate that our proposed models significantly improve diagnostic accuracy,
F1 score, and ROC-AUC compared to conventional approaches. These findings
underscore the potential of integrating transformer architectures with novel
training strategies to advance automated voice pathology detection and
ultimately contribute to more effective healthcare delivery. The code we used
to train and evaluate our models is available at
https://github.com/enkhtogtokh/voicegrpo
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 14:52:57 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Togootogtokh",
"Enkhtogtokh",
""
],
[
"Klasen",
"Christian",
""
]
]
| TITLE: VoiceGRPO: Modern MoE Transformers with Group Relative Policy
Optimization GRPO for AI Voice Health Care Applications on Voice Pathology
Detection
ABSTRACT: This research introduces a novel AI techniques as Mixture-of-Experts
Transformers with Group Relative Policy Optimization (GRPO) for voice health
care applications on voice pathology detection. With the architectural
innovations, we adopt advanced training paradigms inspired by reinforcement
learning, namely Proximal Policy Optimization (PPO) and Group-wise Regularized
Policy Optimization (GRPO), to enhance model stability and performance.
Experiments conducted on a synthetically generated voice pathology dataset
demonstrate that our proposed models significantly improve diagnostic accuracy,
F1 score, and ROC-AUC compared to conventional approaches. These findings
underscore the potential of integrating transformer architectures with novel
training strategies to advance automated voice pathology detection and
ultimately contribute to more effective healthcare delivery. The code we used
to train and evaluate our models is available at
https://github.com/enkhtogtokh/voicegrpo
| new_dataset | 0.959535 |
2503.03803 | Ziwei Liu | Jingkang Yang, Shuai Liu, Hongming Guo, Yuhao Dong, Xiamengwei Zhang,
Sicheng Zhang, Pengyun Wang, Zitang Zhou, Binzhu Xie, Ziyue Wang, Bei Ouyang,
Zhengyu Lin, Marco Cominelli, Zhongang Cai, Yuanhan Zhang, Peiyuan Zhang,
Fangzhou Hong, Joerg Widmer, Francesco Gringoli, Lei Yang, Bo Li, Ziwei Liu | EgoLife: Towards Egocentric Life Assistant | Accepted to CVPR 2025. Project Page: https://egolife-ai.github.io/.
Code: https://github.com/EvolvingLMMs-Lab/EgoLife | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce EgoLife, a project to develop an egocentric life assistant that
accompanies and enhances personal efficiency through AI-powered wearable
glasses. To lay the foundation for this assistant, we conducted a comprehensive
data collection study where six participants lived together for one week,
continuously recording their daily activities - including discussions,
shopping, cooking, socializing, and entertainment - using AI glasses for
multimodal egocentric video capture, along with synchronized third-person-view
video references. This effort resulted in the EgoLife Dataset, a comprehensive
300-hour egocentric, interpersonal, multiview, and multimodal daily life
dataset with intensive annotation. Leveraging this dataset, we introduce
EgoLifeQA, a suite of long-context, life-oriented question-answering tasks
designed to provide meaningful assistance in daily life by addressing practical
questions such as recalling past relevant events, monitoring health habits, and
offering personalized recommendations. To address the key technical challenges
of (1) developing robust visual-audio models for egocentric data, (2) enabling
identity recognition, and (3) facilitating long-context question answering over
extensive temporal information, we introduce EgoButler, an integrated system
comprising EgoGPT and EgoRAG. EgoGPT is an omni-modal model trained on
egocentric datasets, achieving state-of-the-art performance on egocentric video
understanding. EgoRAG is a retrieval-based component that supports answering
ultra-long-context questions. Our experimental studies verify their working
mechanisms and reveal critical factors and bottlenecks, guiding future
improvements. By releasing our datasets, models, and benchmarks, we aim to
stimulate further research in egocentric AI assistants.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 18:54:16 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Yang",
"Jingkang",
""
],
[
"Liu",
"Shuai",
""
],
[
"Guo",
"Hongming",
""
],
[
"Dong",
"Yuhao",
""
],
[
"Zhang",
"Xiamengwei",
""
],
[
"Zhang",
"Sicheng",
""
],
[
"Wang",
"Pengyun",
""
],
[
"Zhou",
"Zitang",
""
],
[
"Xie",
"Binzhu",
""
],
[
"Wang",
"Ziyue",
""
],
[
"Ouyang",
"Bei",
""
],
[
"Lin",
"Zhengyu",
""
],
[
"Cominelli",
"Marco",
""
],
[
"Cai",
"Zhongang",
""
],
[
"Zhang",
"Yuanhan",
""
],
[
"Zhang",
"Peiyuan",
""
],
[
"Hong",
"Fangzhou",
""
],
[
"Widmer",
"Joerg",
""
],
[
"Gringoli",
"Francesco",
""
],
[
"Yang",
"Lei",
""
],
[
"Li",
"Bo",
""
],
[
"Liu",
"Ziwei",
""
]
]
| TITLE: EgoLife: Towards Egocentric Life Assistant
ABSTRACT: We introduce EgoLife, a project to develop an egocentric life assistant that
accompanies and enhances personal efficiency through AI-powered wearable
glasses. To lay the foundation for this assistant, we conducted a comprehensive
data collection study where six participants lived together for one week,
continuously recording their daily activities - including discussions,
shopping, cooking, socializing, and entertainment - using AI glasses for
multimodal egocentric video capture, along with synchronized third-person-view
video references. This effort resulted in the EgoLife Dataset, a comprehensive
300-hour egocentric, interpersonal, multiview, and multimodal daily life
dataset with intensive annotation. Leveraging this dataset, we introduce
EgoLifeQA, a suite of long-context, life-oriented question-answering tasks
designed to provide meaningful assistance in daily life by addressing practical
questions such as recalling past relevant events, monitoring health habits, and
offering personalized recommendations. To address the key technical challenges
of (1) developing robust visual-audio models for egocentric data, (2) enabling
identity recognition, and (3) facilitating long-context question answering over
extensive temporal information, we introduce EgoButler, an integrated system
comprising EgoGPT and EgoRAG. EgoGPT is an omni-modal model trained on
egocentric datasets, achieving state-of-the-art performance on egocentric video
understanding. EgoRAG is a retrieval-based component that supports answering
ultra-long-context questions. Our experimental studies verify their working
mechanisms and reveal critical factors and bottlenecks, guiding future
improvements. By releasing our datasets, models, and benchmarks, we aim to
stimulate further research in egocentric AI assistants.
| new_dataset | 0.569194 |
2503.03826 | Jesse Cresswell | Raunaq Suri, Ilan Gofman, Guangwei Yu, Jesse C. Cresswell | Zero-Execution Retrieval-Augmented Configuration Tuning of Spark
Applications | Code and datasets available at
https://github.com/layer6ai-labs/spark-retrieval-tuning | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large-scale data processing is increasingly done using distributed computing
frameworks like Apache Spark, which have a considerable number of configurable
parameters that affect runtime performance. For optimal performance, these
parameters must be tuned to the specific job being run. Tuning commonly
requires multiple executions to collect runtime information for updating
parameters. This is infeasible for ad hoc queries that are run once or
infrequently. Zero-execution tuning, where parameters are automatically set
before a job's first run, can provide significant savings for all types of
applications, but is more challenging since runtime information is not
available. In this work, we propose a novel method for zero-execution tuning of
Spark configurations based on retrieval. Our method achieves 93.3% of the
runtime improvement of state-of-the-art one-execution optimization, entirely
avoiding the slow initial execution using default settings. The shift to
zero-execution tuning results in a lower cumulative runtime over the first 140
runs, and provides the largest benefit for ad hoc and analytical queries which
only need to be executed once. We release the largest and most comprehensive
suite of Spark query datasets, optimal configurations, and runtime information,
which will promote future development of zero-execution tuning methods.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 19:00:05 GMT"
}
]
| 2025-03-07T00:00:00 | [
[
"Suri",
"Raunaq",
""
],
[
"Gofman",
"Ilan",
""
],
[
"Yu",
"Guangwei",
""
],
[
"Cresswell",
"Jesse C.",
""
]
]
| TITLE: Zero-Execution Retrieval-Augmented Configuration Tuning of Spark
Applications
ABSTRACT: Large-scale data processing is increasingly done using distributed computing
frameworks like Apache Spark, which have a considerable number of configurable
parameters that affect runtime performance. For optimal performance, these
parameters must be tuned to the specific job being run. Tuning commonly
requires multiple executions to collect runtime information for updating
parameters. This is infeasible for ad hoc queries that are run once or
infrequently. Zero-execution tuning, where parameters are automatically set
before a job's first run, can provide significant savings for all types of
applications, but is more challenging since runtime information is not
available. In this work, we propose a novel method for zero-execution tuning of
Spark configurations based on retrieval. Our method achieves 93.3% of the
runtime improvement of state-of-the-art one-execution optimization, entirely
avoiding the slow initial execution using default settings. The shift to
zero-execution tuning results in a lower cumulative runtime over the first 140
runs, and provides the largest benefit for ad hoc and analytical queries which
only need to be executed once. We release the largest and most comprehensive
suite of Spark query datasets, optimal configurations, and runtime information,
which will promote future development of zero-execution tuning methods.
| no_new_dataset | 0.843895 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.