Search is not available for this dataset
id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence | prompt
string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2503.13053 | Nassim Ali Ousalah | Nassim Ali Ousalah, Anis Kacem, Enjie Ghorbel, Emmanuel Koumandakis
and Djamila Aouada | Uncertainty-Aware Knowledge Distillation for Compact and Efficient 6DoF
Pose Estimation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Compact and efficient 6DoF object pose estimation is crucial in applications
such as robotics, augmented reality, and space autonomous navigation systems,
where lightweight models are critical for real-time accurate performance. This
paper introduces a novel uncertainty-aware end-to-end Knowledge Distillation
(KD) framework focused on keypoint-based 6DoF pose estimation. Keypoints
predicted by a large teacher model exhibit varying levels of uncertainty that
can be exploited within the distillation process to enhance the accuracy of the
student model while ensuring its compactness. To this end, we propose a
distillation strategy that aligns the student and teacher predictions by
adjusting the knowledge transfer based on the uncertainty associated with each
teacher keypoint prediction. Additionally, the proposed KD leverages this
uncertainty-aware alignment of keypoints to transfer the knowledge at key
locations of their respective feature maps. Experiments on the widely-used
LINEMOD benchmark demonstrate the effectiveness of our method, achieving
superior 6DoF object pose estimation with lightweight models compared to
state-of-the-art approaches. Further validation on the SPEED+ dataset for
spacecraft pose estimation highlights the robustness of our approach under
diverse 6DoF pose estimation scenarios.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 10:56:30 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Ousalah",
"Nassim Ali",
""
],
[
"Kacem",
"Anis",
""
],
[
"Ghorbel",
"Enjie",
""
],
[
"Koumandakis",
"Emmanuel",
""
],
[
"Aouada",
"Djamila",
""
]
] | TITLE: Uncertainty-Aware Knowledge Distillation for Compact and Efficient 6DoF
Pose Estimation
ABSTRACT: Compact and efficient 6DoF object pose estimation is crucial in applications
such as robotics, augmented reality, and space autonomous navigation systems,
where lightweight models are critical for real-time accurate performance. This
paper introduces a novel uncertainty-aware end-to-end Knowledge Distillation
(KD) framework focused on keypoint-based 6DoF pose estimation. Keypoints
predicted by a large teacher model exhibit varying levels of uncertainty that
can be exploited within the distillation process to enhance the accuracy of the
student model while ensuring its compactness. To this end, we propose a
distillation strategy that aligns the student and teacher predictions by
adjusting the knowledge transfer based on the uncertainty associated with each
teacher keypoint prediction. Additionally, the proposed KD leverages this
uncertainty-aware alignment of keypoints to transfer the knowledge at key
locations of their respective feature maps. Experiments on the widely-used
LINEMOD benchmark demonstrate the effectiveness of our method, achieving
superior 6DoF object pose estimation with lightweight models compared to
state-of-the-art approaches. Further validation on the SPEED+ dataset for
spacecraft pose estimation highlights the robustness of our approach under
diverse 6DoF pose estimation scenarios.
|
2503.13055 | Yu-Hong Shen | Yu-Hong Shen, Chuan-Yu Wu, Yi-Ru Yang, Yen-Ling Tai, Yi-Ting Chen | Mitigating Cross-Modal Distraction and Ensuring Geometric Feasibility
via Affordance-Guided, Self-Consistent MLLMs for Food Preparation Task
Planning | null | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | We study Multimodal Large Language Models (MLLMs) with in-context learning
for food preparation task planning. In this context, we identify two key
challenges: cross-modal distraction and geometric feasibility. Cross-modal
distraction occurs when the inclusion of visual input degrades the reasoning
performance of a MLLM. Geometric feasibility refers to the ability of MLLMs to
ensure that the selected skills are physically executable in the environment.
To address these issues, we adapt Chain of Thought (CoT) with Self-Consistency
to mitigate reasoning loss from cross-modal distractions and use affordance
predictor as skill preconditions to guide MLLM on geometric feasibility. We
construct a dataset to evaluate the ability of MLLMs on quantity estimation,
reachability analysis, relative positioning and collision avoidance. We
conducted a detailed evaluation to identify issues among different baselines
and analyze the reasons for improvement, providing insights into each approach.
Our method reaches a success rate of 76.7% on the entire dataset, showing a
substantial improvement over the CoT baseline at 36.7%.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:01:02 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Shen",
"Yu-Hong",
""
],
[
"Wu",
"Chuan-Yu",
""
],
[
"Yang",
"Yi-Ru",
""
],
[
"Tai",
"Yen-Ling",
""
],
[
"Chen",
"Yi-Ting",
""
]
] | TITLE: Mitigating Cross-Modal Distraction and Ensuring Geometric Feasibility
via Affordance-Guided, Self-Consistent MLLMs for Food Preparation Task
Planning
ABSTRACT: We study Multimodal Large Language Models (MLLMs) with in-context learning
for food preparation task planning. In this context, we identify two key
challenges: cross-modal distraction and geometric feasibility. Cross-modal
distraction occurs when the inclusion of visual input degrades the reasoning
performance of a MLLM. Geometric feasibility refers to the ability of MLLMs to
ensure that the selected skills are physically executable in the environment.
To address these issues, we adapt Chain of Thought (CoT) with Self-Consistency
to mitigate reasoning loss from cross-modal distractions and use affordance
predictor as skill preconditions to guide MLLM on geometric feasibility. We
construct a dataset to evaluate the ability of MLLMs on quantity estimation,
reachability analysis, relative positioning and collision avoidance. We
conducted a detailed evaluation to identify issues among different baselines
and analyze the reasons for improvement, providing insights into each approach.
Our method reaches a success rate of 76.7% on the entire dataset, showing a
substantial improvement over the CoT baseline at 36.7%.
|
2503.13057 | Robin Zbinden | Robin Zbinden, Nina van Tiel, Gencer Sumbul, Chiara Vanalli, Benjamin
Kellenberger, Devis Tuia | MaskSDM with Shapley values to improve flexibility, robustness, and
explainability in species distribution modeling | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Species Distribution Models (SDMs) play a vital role in biodiversity
research, conservation planning, and ecological niche modeling by predicting
species distributions based on environmental conditions. The selection of
predictors is crucial, strongly impacting both model accuracy and how well the
predictions reflect ecological patterns. To ensure meaningful insights, input
variables must be carefully chosen to match the study objectives and the
ecological requirements of the target species. However, existing SDMs,
including both traditional and deep learning-based approaches, often lack key
capabilities for variable selection: (i) flexibility to choose relevant
predictors at inference without retraining; (ii) robustness to handle missing
predictor values without compromising accuracy; and (iii) explainability to
interpret and accurately quantify each predictor's contribution. To overcome
these limitations, we introduce MaskSDM, a novel deep learning-based SDM that
enables flexible predictor selection by employing a masked training strategy.
This approach allows the model to make predictions with arbitrary subsets of
input variables while remaining robust to missing data. It also provides a
clearer understanding of how adding or removing a given predictor affects model
performance and predictions. Additionally, MaskSDM leverages Shapley values for
precise predictor contribution assessments, improving upon traditional
approximations. We evaluate MaskSDM on the global sPlotOpen dataset, modeling
the distributions of 12,738 plant species. Our results show that MaskSDM
outperforms imputation-based methods and approximates models trained on
specific subsets of variables. These findings underscore MaskSDM's potential to
increase the applicability and adoption of SDMs, laying the groundwork for
developing foundation models in SDMs that can be readily applied to diverse
ecological applications.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:02:28 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zbinden",
"Robin",
""
],
[
"van Tiel",
"Nina",
""
],
[
"Sumbul",
"Gencer",
""
],
[
"Vanalli",
"Chiara",
""
],
[
"Kellenberger",
"Benjamin",
""
],
[
"Tuia",
"Devis",
""
]
] | TITLE: MaskSDM with Shapley values to improve flexibility, robustness, and
explainability in species distribution modeling
ABSTRACT: Species Distribution Models (SDMs) play a vital role in biodiversity
research, conservation planning, and ecological niche modeling by predicting
species distributions based on environmental conditions. The selection of
predictors is crucial, strongly impacting both model accuracy and how well the
predictions reflect ecological patterns. To ensure meaningful insights, input
variables must be carefully chosen to match the study objectives and the
ecological requirements of the target species. However, existing SDMs,
including both traditional and deep learning-based approaches, often lack key
capabilities for variable selection: (i) flexibility to choose relevant
predictors at inference without retraining; (ii) robustness to handle missing
predictor values without compromising accuracy; and (iii) explainability to
interpret and accurately quantify each predictor's contribution. To overcome
these limitations, we introduce MaskSDM, a novel deep learning-based SDM that
enables flexible predictor selection by employing a masked training strategy.
This approach allows the model to make predictions with arbitrary subsets of
input variables while remaining robust to missing data. It also provides a
clearer understanding of how adding or removing a given predictor affects model
performance and predictions. Additionally, MaskSDM leverages Shapley values for
precise predictor contribution assessments, improving upon traditional
approximations. We evaluate MaskSDM on the global sPlotOpen dataset, modeling
the distributions of 12,738 plant species. Our results show that MaskSDM
outperforms imputation-based methods and approximates models trained on
specific subsets of variables. These findings underscore MaskSDM's potential to
increase the applicability and adoption of SDMs, laying the groundwork for
developing foundation models in SDMs that can be readily applied to diverse
ecological applications.
|
2503.13058 | Zeyi Huang Mr | Zeyi Huang, Utkarsh Ojha, Yuyang Ji, Donghyun Lee, Yong Jae Lee | Do Vision Models Develop Human-Like Progressive Difficulty
Understanding? | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | When a human undertakes a test, their responses likely follow a pattern: if
they answered an easy question $(2 \times 3)$ incorrectly, they would likely
answer a more difficult one $(2 \times 3 \times 4)$ incorrectly; and if they
answered a difficult question correctly, they would likely answer the easy one
correctly. Anything else hints at memorization. Do current visual recognition
models exhibit a similarly structured learning capacity? In this work, we
consider the task of image classification and study if those models' responses
follow that pattern. Since real images aren't labeled with difficulty, we first
create a dataset of 100 categories, 10 attributes, and 3 difficulty levels
using recent generative models: for each category (e.g., dog) and attribute
(e.g., occlusion), we generate images of increasing difficulty (e.g., a dog
without occlusion, a dog only partly visible). We find that most of the models
do in fact behave similarly to the aforementioned pattern around 80-90% of the
time. Using this property, we then explore a new way to evaluate those models.
Instead of testing the model on every possible test image, we create an
adaptive test akin to GRE, in which the model's performance on the current
round of images determines the test images in the next round. This allows the
model to skip over questions too easy/hard for itself, and helps us get its
overall performance in fewer steps.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:02:53 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Huang",
"Zeyi",
""
],
[
"Ojha",
"Utkarsh",
""
],
[
"Ji",
"Yuyang",
""
],
[
"Lee",
"Donghyun",
""
],
[
"Lee",
"Yong Jae",
""
]
] | TITLE: Do Vision Models Develop Human-Like Progressive Difficulty
Understanding?
ABSTRACT: When a human undertakes a test, their responses likely follow a pattern: if
they answered an easy question $(2 \times 3)$ incorrectly, they would likely
answer a more difficult one $(2 \times 3 \times 4)$ incorrectly; and if they
answered a difficult question correctly, they would likely answer the easy one
correctly. Anything else hints at memorization. Do current visual recognition
models exhibit a similarly structured learning capacity? In this work, we
consider the task of image classification and study if those models' responses
follow that pattern. Since real images aren't labeled with difficulty, we first
create a dataset of 100 categories, 10 attributes, and 3 difficulty levels
using recent generative models: for each category (e.g., dog) and attribute
(e.g., occlusion), we generate images of increasing difficulty (e.g., a dog
without occlusion, a dog only partly visible). We find that most of the models
do in fact behave similarly to the aforementioned pattern around 80-90% of the
time. Using this property, we then explore a new way to evaluate those models.
Instead of testing the model on every possible test image, we create an
adaptive test akin to GRE, in which the model's performance on the current
round of images determines the test images in the next round. This allows the
model to skip over questions too easy/hard for itself, and helps us get its
overall performance in fewer steps.
|
2503.13063 | Zheng Wang | Zheng Wang, Zihui Wang, Zheng Wang, Xiaoliang Fan, Cheng Wang | Federated Learning with Domain Shift Eraser | Accepted by CVPR2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Federated learning (FL) is emerging as a promising technique for
collaborative learning without local data leaving their devices. However,
clients' data originating from diverse domains may degrade model performance
due to domain shifts, preventing the model from learning consistent
representation space. In this paper, we propose a novel FL framework, Federated
Domain Shift Eraser (FDSE), to improve model performance by differently erasing
each client's domain skew and enhancing their consensus. First, we formulate
the model forward passing as an iterative deskewing process that extracts and
then deskews features alternatively. This is efficiently achieved by
decomposing each original layer in the neural network into a Domain-agnostic
Feature Extractor (DFE) and a Domain-specific Skew Eraser (DSE). Then, a
regularization term is applied to promise the effectiveness of feature
deskewing by pulling local statistics of DSE's outputs close to the globally
consistent ones. Finally, DFE modules are fairly aggregated and broadcast to
all the clients to maximize their consensus, and DSE modules are personalized
for each client via similarity-aware aggregation to erase their domain skew
differently. Comprehensive experiments were conducted on three datasets to
confirm the advantages of our method in terms of accuracy, efficiency, and
generalizability.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:10:31 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Wang",
"Zheng",
""
],
[
"Wang",
"Zihui",
""
],
[
"Wang",
"Zheng",
""
],
[
"Fan",
"Xiaoliang",
""
],
[
"Wang",
"Cheng",
""
]
] | TITLE: Federated Learning with Domain Shift Eraser
ABSTRACT: Federated learning (FL) is emerging as a promising technique for
collaborative learning without local data leaving their devices. However,
clients' data originating from diverse domains may degrade model performance
due to domain shifts, preventing the model from learning consistent
representation space. In this paper, we propose a novel FL framework, Federated
Domain Shift Eraser (FDSE), to improve model performance by differently erasing
each client's domain skew and enhancing their consensus. First, we formulate
the model forward passing as an iterative deskewing process that extracts and
then deskews features alternatively. This is efficiently achieved by
decomposing each original layer in the neural network into a Domain-agnostic
Feature Extractor (DFE) and a Domain-specific Skew Eraser (DSE). Then, a
regularization term is applied to promise the effectiveness of feature
deskewing by pulling local statistics of DSE's outputs close to the globally
consistent ones. Finally, DFE modules are fairly aggregated and broadcast to
all the clients to maximize their consensus, and DSE modules are personalized
for each client via similarity-aware aggregation to erase their domain skew
differently. Comprehensive experiments were conducted on three datasets to
confirm the advantages of our method in terms of accuracy, efficiency, and
generalizability.
|
2503.13068 | Henghui Du | Henghui Du, Guangyao Li, Chang Zhou, Chunjie Zhang, Alan Zhao, Di Hu | Crab: A Unified Audio-Visual Scene Understanding Model with Explicit
Cooperation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In recent years, numerous tasks have been proposed to encourage model to
develop specified capability in understanding audio-visual scene, primarily
categorized into temporal localization, spatial localization, spatio-temporal
reasoning, and pixel-level understanding. Instead, human possesses a unified
understanding ability for diversified tasks. Therefore, designing an
audio-visual model with general capability to unify these tasks is of great
value. However, simply joint training for all tasks can lead to interference
due to the heterogeneity of audiovisual data and complex relationship among
tasks. We argue that this problem can be solved through explicit cooperation
among tasks. To achieve this goal, we propose a unified learning method which
achieves explicit inter-task cooperation from both the perspectives of data and
model thoroughly. Specifically, considering the labels of existing datasets are
simple words, we carefully refine these datasets and construct an Audio-Visual
Unified Instruction-tuning dataset with Explicit reasoning process (AV-UIE),
which clarifies the cooperative relationship among tasks. Subsequently, to
facilitate concrete cooperation in learning stage, an interaction-aware LoRA
structure with multiple LoRA heads is designed to learn different aspects of
audiovisual data interaction. By unifying the explicit cooperation across the
data and model aspect, our method not only surpasses existing unified
audio-visual model on multiple tasks, but also outperforms most specialized
models for certain tasks. Furthermore, we also visualize the process of
explicit cooperation and surprisingly find that each LoRA head has certain
audio-visual understanding ability. Code and dataset:
https://github.com/GeWu-Lab/Crab
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:19:03 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Du",
"Henghui",
""
],
[
"Li",
"Guangyao",
""
],
[
"Zhou",
"Chang",
""
],
[
"Zhang",
"Chunjie",
""
],
[
"Zhao",
"Alan",
""
],
[
"Hu",
"Di",
""
]
] | TITLE: Crab: A Unified Audio-Visual Scene Understanding Model with Explicit
Cooperation
ABSTRACT: In recent years, numerous tasks have been proposed to encourage model to
develop specified capability in understanding audio-visual scene, primarily
categorized into temporal localization, spatial localization, spatio-temporal
reasoning, and pixel-level understanding. Instead, human possesses a unified
understanding ability for diversified tasks. Therefore, designing an
audio-visual model with general capability to unify these tasks is of great
value. However, simply joint training for all tasks can lead to interference
due to the heterogeneity of audiovisual data and complex relationship among
tasks. We argue that this problem can be solved through explicit cooperation
among tasks. To achieve this goal, we propose a unified learning method which
achieves explicit inter-task cooperation from both the perspectives of data and
model thoroughly. Specifically, considering the labels of existing datasets are
simple words, we carefully refine these datasets and construct an Audio-Visual
Unified Instruction-tuning dataset with Explicit reasoning process (AV-UIE),
which clarifies the cooperative relationship among tasks. Subsequently, to
facilitate concrete cooperation in learning stage, an interaction-aware LoRA
structure with multiple LoRA heads is designed to learn different aspects of
audiovisual data interaction. By unifying the explicit cooperation across the
data and model aspect, our method not only surpasses existing unified
audio-visual model on multiple tasks, but also outperforms most specialized
models for certain tasks. Furthermore, we also visualize the process of
explicit cooperation and surprisingly find that each LoRA head has certain
audio-visual understanding ability. Code and dataset:
https://github.com/GeWu-Lab/Crab
|
2503.13073 | Zhicheng Zhao | Zhicheng Zhao and Jinquan Yan and Chenglong Li and Xiao Wang and Jin
Tang | DehazeMamba: SAR-guided Optical Remote Sensing Image Dehazing with
Adaptive State Space Model | null | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Optical remote sensing image dehazing presents significant challenges due to
its extensive spatial scale and highly non-uniform haze distribution, which
traditional single-image dehazing methods struggle to address effectively.
While Synthetic Aperture Radar (SAR) imagery offers inherently haze-free
reference information for large-scale scenes, existing SAR-guided dehazing
approaches face two critical limitations: the integration of SAR information
often diminishes the quality of haze-free regions, and the instability of
feature quality further exacerbates cross-modal domain shift. To overcome these
challenges, we introduce DehazeMamba, a novel SAR-guided dehazing network built
on a progressive haze decoupling fusion strategy. Our approach incorporates two
key innovations: a Haze Perception and Decoupling Module (HPDM) that
dynamically identifies haze-affected regions through optical-SAR difference
analysis, and a Progressive Fusion Module (PFM) that mitigates domain shift
through a two-stage fusion process based on feature quality assessment. To
facilitate research in this domain, we present MRSHaze, a large-scale benchmark
dataset comprising 8,000 pairs of temporally synchronized, precisely
geo-registered SAR-optical images with high resolution and diverse haze
conditions. Extensive experiments demonstrate that DehazeMamba significantly
outperforms state-of-the-art methods, achieving a 0.73 dB improvement in PSNR
and substantial enhancements in downstream tasks such as semantic segmentation.
The dataset is available at
https://github.com/mmic-lcl/Datasets-and-benchmark-code.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:25:05 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zhao",
"Zhicheng",
""
],
[
"Yan",
"Jinquan",
""
],
[
"Li",
"Chenglong",
""
],
[
"Wang",
"Xiao",
""
],
[
"Tang",
"Jin",
""
]
] | TITLE: DehazeMamba: SAR-guided Optical Remote Sensing Image Dehazing with
Adaptive State Space Model
ABSTRACT: Optical remote sensing image dehazing presents significant challenges due to
its extensive spatial scale and highly non-uniform haze distribution, which
traditional single-image dehazing methods struggle to address effectively.
While Synthetic Aperture Radar (SAR) imagery offers inherently haze-free
reference information for large-scale scenes, existing SAR-guided dehazing
approaches face two critical limitations: the integration of SAR information
often diminishes the quality of haze-free regions, and the instability of
feature quality further exacerbates cross-modal domain shift. To overcome these
challenges, we introduce DehazeMamba, a novel SAR-guided dehazing network built
on a progressive haze decoupling fusion strategy. Our approach incorporates two
key innovations: a Haze Perception and Decoupling Module (HPDM) that
dynamically identifies haze-affected regions through optical-SAR difference
analysis, and a Progressive Fusion Module (PFM) that mitigates domain shift
through a two-stage fusion process based on feature quality assessment. To
facilitate research in this domain, we present MRSHaze, a large-scale benchmark
dataset comprising 8,000 pairs of temporally synchronized, precisely
geo-registered SAR-optical images with high resolution and diverse haze
conditions. Extensive experiments demonstrate that DehazeMamba significantly
outperforms state-of-the-art methods, achieving a 0.73 dB improvement in PSNR
and substantial enhancements in downstream tasks such as semantic segmentation.
The dataset is available at
https://github.com/mmic-lcl/Datasets-and-benchmark-code.
|
2503.13082 | Runyu Jiao | Runyu Jiao, Alice Fasoli, Francesco Giuliari, Matteo Bortolon, Sergio
Povoli, Guofeng Mei, Yiming Wang, Fabio Poiesi | Free-form language-based robotic reasoning and grasping | Project website: https://tev-fbk.github.io/FreeGrasp/ | null | null | null | cs.RO cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Performing robotic grasping from a cluttered bin based on human instructions
is a challenging task, as it requires understanding both the nuances of
free-form language and the spatial relationships between objects.
Vision-Language Models (VLMs) trained on web-scale data, such as GPT-4o, have
demonstrated remarkable reasoning capabilities across both text and images. But
can they truly be used for this task in a zero-shot setting? And what are their
limitations? In this paper, we explore these research questions via the
free-form language-based robotic grasping task, and propose a novel method,
FreeGrasp, leveraging the pre-trained VLMs' world knowledge to reason about
human instructions and object spatial arrangements. Our method detects all
objects as keypoints and uses these keypoints to annotate marks on images,
aiming to facilitate GPT-4o's zero-shot spatial reasoning. This allows our
method to determine whether a requested object is directly graspable or if
other objects must be grasped and removed first. Since no existing dataset is
specifically designed for this task, we introduce a synthetic dataset
FreeGraspData by extending the MetaGraspNetV2 dataset with human-annotated
instructions and ground-truth grasping sequences. We conduct extensive analyses
with both FreeGraspData and real-world validation with a gripper-equipped
robotic arm, demonstrating state-of-the-art performance in grasp reasoning and
execution. Project website: https://tev-fbk.github.io/FreeGrasp/.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:41:16 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Jiao",
"Runyu",
""
],
[
"Fasoli",
"Alice",
""
],
[
"Giuliari",
"Francesco",
""
],
[
"Bortolon",
"Matteo",
""
],
[
"Povoli",
"Sergio",
""
],
[
"Mei",
"Guofeng",
""
],
[
"Wang",
"Yiming",
""
],
[
"Poiesi",
"Fabio",
""
]
] | TITLE: Free-form language-based robotic reasoning and grasping
ABSTRACT: Performing robotic grasping from a cluttered bin based on human instructions
is a challenging task, as it requires understanding both the nuances of
free-form language and the spatial relationships between objects.
Vision-Language Models (VLMs) trained on web-scale data, such as GPT-4o, have
demonstrated remarkable reasoning capabilities across both text and images. But
can they truly be used for this task in a zero-shot setting? And what are their
limitations? In this paper, we explore these research questions via the
free-form language-based robotic grasping task, and propose a novel method,
FreeGrasp, leveraging the pre-trained VLMs' world knowledge to reason about
human instructions and object spatial arrangements. Our method detects all
objects as keypoints and uses these keypoints to annotate marks on images,
aiming to facilitate GPT-4o's zero-shot spatial reasoning. This allows our
method to determine whether a requested object is directly graspable or if
other objects must be grasped and removed first. Since no existing dataset is
specifically designed for this task, we introduce a synthetic dataset
FreeGraspData by extending the MetaGraspNetV2 dataset with human-annotated
instructions and ground-truth grasping sequences. We conduct extensive analyses
with both FreeGraspData and real-world validation with a gripper-equipped
robotic arm, demonstrating state-of-the-art performance in grasp reasoning and
execution. Project website: https://tev-fbk.github.io/FreeGrasp/.
|
2503.13086 | Yiwei Xu | Yiwei Xu, Yifei Yu, Wentian Gan, Tengfei Wang, Zongqian Zhan, Hao
Cheng and Xin Wang | Gaussian On-the-Fly Splatting: A Progressive Framework for Robust Near
Real-Time 3DGS Optimization | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D Gaussian Splatting (3DGS) achieves high-fidelity rendering with fast
real-time performance, but existing methods rely on offline training after full
Structure-from-Motion (SfM) processing. In contrast, this work introduces
On-the-Fly GS, a progressive framework enabling near real-time 3DGS
optimization during image capture. As each image arrives, its pose and sparse
points are updated via on-the-fly SfM, and newly optimized Gaussians are
immediately integrated into the 3DGS field. We propose a progressive local
optimization strategy to prioritize new images and their neighbors by their
corresponding overlapping relationship, allowing the new image and its
overlapping images to get more training. To further stabilize training across
old and new images, an adaptive learning rate schedule balances the iterations
and the learning rate. Moreover, to maintain overall quality of the 3DGS field,
an efficient global optimization scheme prevents overfitting to the newly added
images. Experiments on multiple benchmark datasets show that our On-the-Fly GS
reduces training time significantly, optimizing each new image in seconds with
minimal rendering loss, offering the first practical step toward rapid,
progressive 3DGS reconstruction.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:47:58 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Xu",
"Yiwei",
""
],
[
"Yu",
"Yifei",
""
],
[
"Gan",
"Wentian",
""
],
[
"Wang",
"Tengfei",
""
],
[
"Zhan",
"Zongqian",
""
],
[
"Cheng",
"Hao",
""
],
[
"Wang",
"Xin",
""
]
] | TITLE: Gaussian On-the-Fly Splatting: A Progressive Framework for Robust Near
Real-Time 3DGS Optimization
ABSTRACT: 3D Gaussian Splatting (3DGS) achieves high-fidelity rendering with fast
real-time performance, but existing methods rely on offline training after full
Structure-from-Motion (SfM) processing. In contrast, this work introduces
On-the-Fly GS, a progressive framework enabling near real-time 3DGS
optimization during image capture. As each image arrives, its pose and sparse
points are updated via on-the-fly SfM, and newly optimized Gaussians are
immediately integrated into the 3DGS field. We propose a progressive local
optimization strategy to prioritize new images and their neighbors by their
corresponding overlapping relationship, allowing the new image and its
overlapping images to get more training. To further stabilize training across
old and new images, an adaptive learning rate schedule balances the iterations
and the learning rate. Moreover, to maintain overall quality of the 3DGS field,
an efficient global optimization scheme prevents overfitting to the newly added
images. Experiments on multiple benchmark datasets show that our On-the-Fly GS
reduces training time significantly, optimizing each new image in seconds with
minimal rendering loss, offering the first practical step toward rapid,
progressive 3DGS reconstruction.
|
2503.13090 | Tom\'a\v{s} Pivo\v{n}ka | V\'aclav Truhla\v{r}\'ik, Tom\'a\v{s} Pivo\v{n}ka, Michal Kasarda,
Libor P\v{r}eu\v{c}il | Multi-Platform Teach-and-Repeat Navigation by Visual Place Recognition
Based on Deep-Learned Local Features | 6 pages, 5 figures | null | null | null | cs.RO cs.CV | http://creativecommons.org/licenses/by/4.0/ | Uniform and variable environments still remain a challenge for stable visual
localization and mapping in mobile robot navigation. One of the possible
approaches suitable for such environments is appearance-based teach-and-repeat
navigation, relying on simplified localization and reactive robot motion
control - all without a need for standard mapping. This work brings an
innovative solution to such a system based on visual place recognition
techniques. Here, the major contributions stand in the employment of a new
visual place recognition technique, a novel horizontal shift computation
approach, and a multi-platform system design for applications across various
types of mobile robots. Secondly, a new public dataset for experimental testing
of appearance-based navigation methods is introduced. Moreover, the work also
provides real-world experimental testing and performance comparison of the
introduced navigation system against other state-of-the-art methods. The
results confirm that the new system outperforms existing methods in several
testing scenarios, is capable of operation indoors and outdoors, and exhibits
robustness to day and night scene variations.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:57:41 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Truhlařík",
"Václav",
""
],
[
"Pivoňka",
"Tomáš",
""
],
[
"Kasarda",
"Michal",
""
],
[
"Přeučil",
"Libor",
""
]
] | TITLE: Multi-Platform Teach-and-Repeat Navigation by Visual Place Recognition
Based on Deep-Learned Local Features
ABSTRACT: Uniform and variable environments still remain a challenge for stable visual
localization and mapping in mobile robot navigation. One of the possible
approaches suitable for such environments is appearance-based teach-and-repeat
navigation, relying on simplified localization and reactive robot motion
control - all without a need for standard mapping. This work brings an
innovative solution to such a system based on visual place recognition
techniques. Here, the major contributions stand in the employment of a new
visual place recognition technique, a novel horizontal shift computation
approach, and a multi-platform system design for applications across various
types of mobile robots. Secondly, a new public dataset for experimental testing
of appearance-based navigation methods is introduced. Moreover, the work also
provides real-world experimental testing and performance comparison of the
introduced navigation system against other state-of-the-art methods. The
results confirm that the new system outperforms existing methods in several
testing scenarios, is capable of operation indoors and outdoors, and exhibits
robustness to day and night scene variations.
|
2503.13101 | Babangida Sani | Babangida Sani, Aakansha Soy, Sukairaj Hafiz Imam, Ahmad Mustapha,
Lukman Jibril Aliyu, Idris Abdulmumin, Ibrahim Said Ahmad, Shamsuddeen Hassan
Muhammad | Who Wrote This? Identifying Machine vs Human-Generated Text in Hausa | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The advancement of large language models (LLMs) has allowed them to be
proficient in various tasks, including content generation. However, their
unregulated usage can lead to malicious activities such as plagiarism and
generating and spreading fake news, especially for low-resource languages. Most
existing machine-generated text detectors are trained on high-resource
languages like English, French, etc. In this study, we developed the first
large-scale detector that can distinguish between human- and machine-generated
content in Hausa. We scrapped seven Hausa-language media outlets for the
human-generated text and the Gemini-2.0 flash model to automatically generate
the corresponding Hausa-language articles based on the human-generated article
headlines. We fine-tuned four pre-trained Afri-centric models (AfriTeVa,
AfriBERTa, AfroXLMR, and AfroXLMR-76L) on the resulting dataset and assessed
their performance using accuracy and F1-score metrics. AfroXLMR achieved the
highest performance with an accuracy of 99.23% and an F1 score of 99.21%,
demonstrating its effectiveness for Hausa text detection. Our dataset is made
publicly available to enable further research.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:13:37 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Sani",
"Babangida",
""
],
[
"Soy",
"Aakansha",
""
],
[
"Imam",
"Sukairaj Hafiz",
""
],
[
"Mustapha",
"Ahmad",
""
],
[
"Aliyu",
"Lukman Jibril",
""
],
[
"Abdulmumin",
"Idris",
""
],
[
"Ahmad",
"Ibrahim Said",
""
],
[
"Muhammad",
"Shamsuddeen Hassan",
""
]
] | TITLE: Who Wrote This? Identifying Machine vs Human-Generated Text in Hausa
ABSTRACT: The advancement of large language models (LLMs) has allowed them to be
proficient in various tasks, including content generation. However, their
unregulated usage can lead to malicious activities such as plagiarism and
generating and spreading fake news, especially for low-resource languages. Most
existing machine-generated text detectors are trained on high-resource
languages like English, French, etc. In this study, we developed the first
large-scale detector that can distinguish between human- and machine-generated
content in Hausa. We scrapped seven Hausa-language media outlets for the
human-generated text and the Gemini-2.0 flash model to automatically generate
the corresponding Hausa-language articles based on the human-generated article
headlines. We fine-tuned four pre-trained Afri-centric models (AfriTeVa,
AfriBERTa, AfroXLMR, and AfroXLMR-76L) on the resulting dataset and assessed
their performance using accuracy and F1-score metrics. AfroXLMR achieved the
highest performance with an accuracy of 99.23% and an F1 score of 99.21%,
demonstrating its effectiveness for Hausa text detection. Our dataset is made
publicly available to enable further research.
|
2503.13102 | Ekaterina Artemova | Alexander Pugachev, Alena Fenogenova, Vladislav Mikhailov, Ekaterina
Artemova | REPA: Russian Error Types Annotation for Evaluating Text Generation and
Judgment Capabilities | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Recent advances in large language models (LLMs) have introduced the novel
paradigm of using LLMs as judges, where an LLM evaluates and scores the outputs
of another LLM, which often correlates highly with human preferences. However,
the use of LLM-as-a-judge has been primarily studied in English. In this paper,
we evaluate this framework in Russian by introducing the Russian Error tyPes
Annotation dataset (REPA), a dataset of 1k user queries and 2k LLM-generated
responses. Human annotators labeled each response pair expressing their
preferences across ten specific error types, as well as selecting an overall
preference. We rank six generative LLMs across the error types using three
rating systems based on human preferences. We also evaluate responses using
eight LLM judges in zero-shot and few-shot settings. We describe the results of
analyzing the judges and position and length biases. Our findings reveal a
notable gap between LLM judge performance in Russian and English. However,
rankings based on human and LLM preferences show partial alignment, suggesting
that while current LLM judges struggle with fine-grained evaluation in Russian,
there is potential for improvement.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:15:16 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Pugachev",
"Alexander",
""
],
[
"Fenogenova",
"Alena",
""
],
[
"Mikhailov",
"Vladislav",
""
],
[
"Artemova",
"Ekaterina",
""
]
] | TITLE: REPA: Russian Error Types Annotation for Evaluating Text Generation and
Judgment Capabilities
ABSTRACT: Recent advances in large language models (LLMs) have introduced the novel
paradigm of using LLMs as judges, where an LLM evaluates and scores the outputs
of another LLM, which often correlates highly with human preferences. However,
the use of LLM-as-a-judge has been primarily studied in English. In this paper,
we evaluate this framework in Russian by introducing the Russian Error tyPes
Annotation dataset (REPA), a dataset of 1k user queries and 2k LLM-generated
responses. Human annotators labeled each response pair expressing their
preferences across ten specific error types, as well as selecting an overall
preference. We rank six generative LLMs across the error types using three
rating systems based on human preferences. We also evaluate responses using
eight LLM judges in zero-shot and few-shot settings. We describe the results of
analyzing the judges and position and length biases. Our findings reveal a
notable gap between LLM judge performance in Russian and English. However,
rankings based on human and LLM preferences show partial alignment, suggesting
that while current LLM judges struggle with fine-grained evaluation in Russian,
there is potential for improvement.
|
2503.13109 | Kedi Chen | Kedi Chen, Zhikai Lei, Fan Zhang, Yinqi Zhang, Qin Chen, Jie Zhou,
Liang He, Qipeng Guo, Kai Chen, Wei Zhang | Code-Driven Inductive Synthesis: Enhancing Reasoning Abilities of Large
Language Models with Sequences | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large language models make remarkable progress in reasoning capabilities.
Existing works focus mainly on deductive reasoning tasks (e.g., code and math),
while another type of reasoning mode that better aligns with human learning,
inductive reasoning, is not well studied. We attribute the reason to the fact
that obtaining high-quality process supervision data is challenging for
inductive reasoning. Towards this end, we novelly employ number sequences as
the source of inductive reasoning data. We package sequences into algorithmic
problems to find the general term of each sequence through a code solution. In
this way, we can verify whether the code solution holds for any term in the
current sequence, and inject case-based supervision signals by using code unit
tests. We build a sequence synthetic data pipeline and form a training dataset
CodeSeq. Experimental results show that the models tuned with CodeSeq improve
on both code and comprehensive reasoning benchmarks.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:33:26 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Chen",
"Kedi",
""
],
[
"Lei",
"Zhikai",
""
],
[
"Zhang",
"Fan",
""
],
[
"Zhang",
"Yinqi",
""
],
[
"Chen",
"Qin",
""
],
[
"Zhou",
"Jie",
""
],
[
"He",
"Liang",
""
],
[
"Guo",
"Qipeng",
""
],
[
"Chen",
"Kai",
""
],
[
"Zhang",
"Wei",
""
]
] | TITLE: Code-Driven Inductive Synthesis: Enhancing Reasoning Abilities of Large
Language Models with Sequences
ABSTRACT: Large language models make remarkable progress in reasoning capabilities.
Existing works focus mainly on deductive reasoning tasks (e.g., code and math),
while another type of reasoning mode that better aligns with human learning,
inductive reasoning, is not well studied. We attribute the reason to the fact
that obtaining high-quality process supervision data is challenging for
inductive reasoning. Towards this end, we novelly employ number sequences as
the source of inductive reasoning data. We package sequences into algorithmic
problems to find the general term of each sequence through a code solution. In
this way, we can verify whether the code solution holds for any term in the
current sequence, and inject case-based supervision signals by using code unit
tests. We build a sequence synthetic data pipeline and form a training dataset
CodeSeq. Experimental results show that the models tuned with CodeSeq improve
on both code and comprehensive reasoning benchmarks.
|
2503.13110 | Jing Li | Jing Li, Yihang Fu, Falai Chen | DTGBrepGen: A Novel B-rep Generative Model through Decoupling Topology
and Geometry | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Boundary representation (B-rep) of geometric models is a fundamental format
in Computer-Aided Design (CAD). However, automatically generating valid and
high-quality B-rep models remains challenging due to the complex
interdependence between the topology and geometry of the models. Existing
methods tend to prioritize geometric representation while giving insufficient
attention to topological constraints, making it difficult to maintain
structural validity and geometric accuracy. In this paper, we propose
DTGBrepGen, a novel topology-geometry decoupled framework for B-rep generation
that explicitly addresses both aspects. Our approach first generates valid
topological structures through a two-stage process that independently models
edge-face and edge-vertex adjacency relationships. Subsequently, we employ
Transformer-based diffusion models for sequential geometry generation,
progressively generating vertex coordinates, followed by edge geometries and
face geometries which are represented as B-splines. Extensive experiments on
diverse CAD datasets show that DTGBrepGen significantly outperforms existing
methods in both topological validity and geometric accuracy, achieving higher
validity rates and producing more diverse and realistic B-reps. Our code is
publicly available at https://github.com/jinli99/DTGBrepGen.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:34:14 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Li",
"Jing",
""
],
[
"Fu",
"Yihang",
""
],
[
"Chen",
"Falai",
""
]
] | TITLE: DTGBrepGen: A Novel B-rep Generative Model through Decoupling Topology
and Geometry
ABSTRACT: Boundary representation (B-rep) of geometric models is a fundamental format
in Computer-Aided Design (CAD). However, automatically generating valid and
high-quality B-rep models remains challenging due to the complex
interdependence between the topology and geometry of the models. Existing
methods tend to prioritize geometric representation while giving insufficient
attention to topological constraints, making it difficult to maintain
structural validity and geometric accuracy. In this paper, we propose
DTGBrepGen, a novel topology-geometry decoupled framework for B-rep generation
that explicitly addresses both aspects. Our approach first generates valid
topological structures through a two-stage process that independently models
edge-face and edge-vertex adjacency relationships. Subsequently, we employ
Transformer-based diffusion models for sequential geometry generation,
progressively generating vertex coordinates, followed by edge geometries and
face geometries which are represented as B-splines. Extensive experiments on
diverse CAD datasets show that DTGBrepGen significantly outperforms existing
methods in both topological validity and geometric accuracy, achieving higher
validity rates and producing more diverse and realistic B-reps. Our code is
publicly available at https://github.com/jinli99/DTGBrepGen.
|
2503.13111 | Erik Daxberger | Erik Daxberger, Nina Wenzel, David Griffiths, Haiming Gang, Justin
Lazarow, Gefen Kohavi, Kai Kang, Marcin Eichner, Yinfei Yang, Afshin Dehghan,
Peter Grasch | MM-Spatial: Exploring 3D Spatial Understanding in Multimodal LLMs | null | null | null | null | cs.CV cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal large language models (MLLMs) excel at 2D visual understanding but
remain limited in their ability to reason about 3D space. In this work, we
leverage large-scale high-quality 3D scene data with open-set annotations to
introduce 1) a novel supervised fine-tuning dataset and 2) a new evaluation
benchmark, focused on indoor scenes. Our Cubify Anything VQA (CA-VQA) data
covers diverse spatial tasks including spatial relationship prediction, metric
size and distance estimation, and 3D grounding. We show that CA-VQA enables us
to train MM-Spatial, a strong generalist MLLM that also achieves
state-of-the-art performance on 3D spatial understanding benchmarks, including
our own. We show how incorporating metric depth and multi-view inputs (provided
in CA-VQA) can further improve 3D understanding, and demonstrate that data
alone allows our model to achieve depth perception capabilities comparable to
dedicated monocular depth estimation models. We will publish our SFT dataset
and benchmark.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:34:22 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Daxberger",
"Erik",
""
],
[
"Wenzel",
"Nina",
""
],
[
"Griffiths",
"David",
""
],
[
"Gang",
"Haiming",
""
],
[
"Lazarow",
"Justin",
""
],
[
"Kohavi",
"Gefen",
""
],
[
"Kang",
"Kai",
""
],
[
"Eichner",
"Marcin",
""
],
[
"Yang",
"Yinfei",
""
],
[
"Dehghan",
"Afshin",
""
],
[
"Grasch",
"Peter",
""
]
] | TITLE: MM-Spatial: Exploring 3D Spatial Understanding in Multimodal LLMs
ABSTRACT: Multimodal large language models (MLLMs) excel at 2D visual understanding but
remain limited in their ability to reason about 3D space. In this work, we
leverage large-scale high-quality 3D scene data with open-set annotations to
introduce 1) a novel supervised fine-tuning dataset and 2) a new evaluation
benchmark, focused on indoor scenes. Our Cubify Anything VQA (CA-VQA) data
covers diverse spatial tasks including spatial relationship prediction, metric
size and distance estimation, and 3D grounding. We show that CA-VQA enables us
to train MM-Spatial, a strong generalist MLLM that also achieves
state-of-the-art performance on 3D spatial understanding benchmarks, including
our own. We show how incorporating metric depth and multi-view inputs (provided
in CA-VQA) can further improve 3D understanding, and demonstrate that data
alone allows our model to achieve depth perception capabilities comparable to
dedicated monocular depth estimation models. We will publish our SFT dataset
and benchmark.
|
2503.13113 | Arjun Pakrashi | Gabriele Sanguin, Arjun Pakrashi, Marco Viola, Francesco Rinaldi | Exploring the Potential of Bilevel Optimization for Calibrating Neural
Networks | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Handling uncertainty is critical for ensuring reliable decision-making in
intelligent systems. Modern neural networks are known to be poorly calibrated,
resulting in predicted confidence scores that are difficult to use. This
article explores improving confidence estimation and calibration through the
application of bilevel optimization, a framework designed to solve hierarchical
problems with interdependent optimization levels. A self-calibrating bilevel
neural-network training approach is introduced to improve a model's predicted
confidence scores. The effectiveness of the proposed framework is analyzed
using toy datasets, such as Blobs and Spirals, as well as more practical
simulated datasets, such as Blood Alcohol Concentration (BAC). It is compared
with a well-known and widely used calibration strategy, isotonic regression.
The reported experimental results reveal that the proposed bilevel optimization
approach reduces the calibration error while preserving accuracy.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:34:55 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Sanguin",
"Gabriele",
""
],
[
"Pakrashi",
"Arjun",
""
],
[
"Viola",
"Marco",
""
],
[
"Rinaldi",
"Francesco",
""
]
] | TITLE: Exploring the Potential of Bilevel Optimization for Calibrating Neural
Networks
ABSTRACT: Handling uncertainty is critical for ensuring reliable decision-making in
intelligent systems. Modern neural networks are known to be poorly calibrated,
resulting in predicted confidence scores that are difficult to use. This
article explores improving confidence estimation and calibration through the
application of bilevel optimization, a framework designed to solve hierarchical
problems with interdependent optimization levels. A self-calibrating bilevel
neural-network training approach is introduced to improve a model's predicted
confidence scores. The effectiveness of the proposed framework is analyzed
using toy datasets, such as Blobs and Spirals, as well as more practical
simulated datasets, such as Blood Alcohol Concentration (BAC). It is compared
with a well-known and widely used calibration strategy, isotonic regression.
The reported experimental results reveal that the proposed bilevel optimization
approach reduces the calibration error while preserving accuracy.
|
2503.13116 | Zeng Wang | Zeng Wang, Minghao Shao, Mohammed Nabeel, Prithwish Basu Roy, Likhitha
Mankali, Jitendra Bhandari, Ramesh Karri, Ozgur Sinanoglu, Muhammad Shafique,
Johann Knechtel | VeriLeaky: Navigating IP Protection vs Utility in Fine-Tuning for
LLM-Driven Verilog Coding | null | null | null | null | cs.CR cs.AR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) offer significant potential for coding, yet
fine-tuning (FT) with curated data is essential for niche languages like
Verilog. Using proprietary intellectual property (IP) for FT presents a serious
risk, as FT data can be leaked through LLM inference. This leads to a critical
dilemma for design houses: seeking to build externally accessible LLMs offering
competitive Verilog coding, how can they leverage in-house IP to enhance FT
utility while ensuring IP protection?
For the first time in the literature, we study this dilemma. Using LLaMA
3.1-8B, we conduct in-house FT on a baseline Verilog dataset (RTLCoder)
supplemented with our own in-house IP, which is validated through multiple
tape-outs. To rigorously assess IP leakage, we quantify structural similarity
(AST/Dolos) and functional equivalence (Synopsys Formality) between generated
codes and our in-house IP. We show that our IP can indeed be leaked, confirming
the threat. As defense, we evaluate logic locking of Verilog codes (ASSURE).
This offers some level of protection, yet reduces the IP's utility for FT and
degrades the LLM's performance. Our study shows the need for novel strategies
that are both effective and minimally disruptive to FT, an essential effort for
enabling design houses to fully utilize their proprietary IP toward LLM-driven
Verilog coding.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:38:03 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Wang",
"Zeng",
""
],
[
"Shao",
"Minghao",
""
],
[
"Nabeel",
"Mohammed",
""
],
[
"Roy",
"Prithwish Basu",
""
],
[
"Mankali",
"Likhitha",
""
],
[
"Bhandari",
"Jitendra",
""
],
[
"Karri",
"Ramesh",
""
],
[
"Sinanoglu",
"Ozgur",
""
],
[
"Shafique",
"Muhammad",
""
],
[
"Knechtel",
"Johann",
""
]
] | TITLE: VeriLeaky: Navigating IP Protection vs Utility in Fine-Tuning for
LLM-Driven Verilog Coding
ABSTRACT: Large language models (LLMs) offer significant potential for coding, yet
fine-tuning (FT) with curated data is essential for niche languages like
Verilog. Using proprietary intellectual property (IP) for FT presents a serious
risk, as FT data can be leaked through LLM inference. This leads to a critical
dilemma for design houses: seeking to build externally accessible LLMs offering
competitive Verilog coding, how can they leverage in-house IP to enhance FT
utility while ensuring IP protection?
For the first time in the literature, we study this dilemma. Using LLaMA
3.1-8B, we conduct in-house FT on a baseline Verilog dataset (RTLCoder)
supplemented with our own in-house IP, which is validated through multiple
tape-outs. To rigorously assess IP leakage, we quantify structural similarity
(AST/Dolos) and functional equivalence (Synopsys Formality) between generated
codes and our in-house IP. We show that our IP can indeed be leaked, confirming
the threat. As defense, we evaluate logic locking of Verilog codes (ASSURE).
This offers some level of protection, yet reduces the IP's utility for FT and
degrades the LLM's performance. Our study shows the need for novel strategies
that are both effective and minimally disruptive to FT, an essential effort for
enabling design houses to fully utilize their proprietary IP toward LLM-driven
Verilog coding.
|
2503.13120 | Siyuan Fan | Siyuan Fan, Wenke Huang, Xiantao Cai, Bo Du | 3D Human Interaction Generation: A Survey | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D human interaction generation has emerged as a key research area, focusing
on producing dynamic and contextually relevant interactions between humans and
various interactive entities. Recent rapid advancements in 3D model
representation methods, motion capture technologies, and generative models have
laid a solid foundation for the growing interest in this domain. Existing
research in this field can be broadly categorized into three areas: human-scene
interaction, human-object interaction, and human-human interaction. Despite the
rapid advancements in this area, challenges remain due to the need for
naturalness in human motion generation and the accurate interaction between
humans and interactive entities. In this survey, we present a comprehensive
literature review of human interaction generation, which, to the best of our
knowledge, is the first of its kind. We begin by introducing the foundational
technologies, including model representations, motion capture methods, and
generative models. Subsequently, we introduce the approaches proposed for the
three sub-tasks, along with their corresponding datasets and evaluation
metrics. Finally, we discuss potential future research directions in this area
and conclude the survey. Through this survey, we aim to offer a comprehensive
overview of the current advancements in the field, highlight key challenges,
and inspire future research works.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:47:33 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Fan",
"Siyuan",
""
],
[
"Huang",
"Wenke",
""
],
[
"Cai",
"Xiantao",
""
],
[
"Du",
"Bo",
""
]
] | TITLE: 3D Human Interaction Generation: A Survey
ABSTRACT: 3D human interaction generation has emerged as a key research area, focusing
on producing dynamic and contextually relevant interactions between humans and
various interactive entities. Recent rapid advancements in 3D model
representation methods, motion capture technologies, and generative models have
laid a solid foundation for the growing interest in this domain. Existing
research in this field can be broadly categorized into three areas: human-scene
interaction, human-object interaction, and human-human interaction. Despite the
rapid advancements in this area, challenges remain due to the need for
naturalness in human motion generation and the accurate interaction between
humans and interactive entities. In this survey, we present a comprehensive
literature review of human interaction generation, which, to the best of our
knowledge, is the first of its kind. We begin by introducing the foundational
technologies, including model representations, motion capture methods, and
generative models. Subsequently, we introduce the approaches proposed for the
three sub-tasks, along with their corresponding datasets and evaluation
metrics. Finally, we discuss potential future research directions in this area
and conclude the survey. Through this survey, we aim to offer a comprehensive
overview of the current advancements in the field, highlight key challenges,
and inspire future research works.
|
2503.13125 | Pan Liu | Pan Liu | Non-Destructive Detection of Sub-Micron Imperceptible Scratches On Laser
Chips Based On Consistent Texture Entropy Recursive Optimization
Semi-Supervised Network | 11 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Laser chips, the core components of semiconductor lasers, are extensively
utilized in various industries, showing great potential for future application.
Smoothness emitting surfaces are crucial in chip production, as even
imperceptible scratches can significantly degrade performance and lifespan,
thus impeding production efficiency and yield. Therefore, non-destructively
detecting these imperceptible scratches on the emitting surfaces is essential
for enhancing yield and reducing costs. These sub-micron level scratches,
barely visible against the background, are extremely difficult to detect with
conventional methods, compounded by a lack of labeled datasets. To address this
challenge, this paper introduces TexRecNet, a consistent texture entropy
recursive optimization semi-supervised network. The network, based on a
recursive optimization architecture, iteratively improves the detection
accuracy of imperceptible scratch edges, using outputs from previous cycles to
inform subsequent inputs and guide the network's positional encoding. It also
introduces image texture entropy, utilizing a substantial amount of unlabeled
data to expand the training set while maintaining training signal reliability.
Ultimately, by analyzing the inconsistency of the network output sequences
obtained during the recursive process, a semi-supervised training strategy with
recursive consistency constraints is proposed, using outputs from the recursive
process for non-destructive signal augmentation and consistently optimizes the
loss function for efficient end-to-end training. Experimental results show that
this method, utilizing a substantial amount of unsupervised data, achieves
75.6% accuracy and 74.8% recall in detecting imperceptible scratches, an 8.5%
and 33.6% improvement over conventional Unet, enhancing quality control in
laser chips.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:48:48 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Liu",
"Pan",
""
]
] | TITLE: Non-Destructive Detection of Sub-Micron Imperceptible Scratches On Laser
Chips Based On Consistent Texture Entropy Recursive Optimization
Semi-Supervised Network
ABSTRACT: Laser chips, the core components of semiconductor lasers, are extensively
utilized in various industries, showing great potential for future application.
Smoothness emitting surfaces are crucial in chip production, as even
imperceptible scratches can significantly degrade performance and lifespan,
thus impeding production efficiency and yield. Therefore, non-destructively
detecting these imperceptible scratches on the emitting surfaces is essential
for enhancing yield and reducing costs. These sub-micron level scratches,
barely visible against the background, are extremely difficult to detect with
conventional methods, compounded by a lack of labeled datasets. To address this
challenge, this paper introduces TexRecNet, a consistent texture entropy
recursive optimization semi-supervised network. The network, based on a
recursive optimization architecture, iteratively improves the detection
accuracy of imperceptible scratch edges, using outputs from previous cycles to
inform subsequent inputs and guide the network's positional encoding. It also
introduces image texture entropy, utilizing a substantial amount of unlabeled
data to expand the training set while maintaining training signal reliability.
Ultimately, by analyzing the inconsistency of the network output sequences
obtained during the recursive process, a semi-supervised training strategy with
recursive consistency constraints is proposed, using outputs from the recursive
process for non-destructive signal augmentation and consistently optimizes the
loss function for efficient end-to-end training. Experimental results show that
this method, utilizing a substantial amount of unsupervised data, achieves
75.6% accuracy and 74.8% recall in detecting imperceptible scratches, an 8.5%
and 33.6% improvement over conventional Unet, enhancing quality control in
laser chips.
|
2503.13130 | Ling-An Zeng | Ling-An Zeng, Guohong Huang, Yi-Lin Wei, Shengbo Gu, Yu-Ming Tang,
Jingke Meng, Wei-Shi Zheng | ChainHOI: Joint-based Kinematic Chain Modeling for Human-Object
Interaction Generation | Accepted to CVPR 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose ChainHOI, a novel approach for text-driven human-object
interaction (HOI) generation that explicitly models interactions at both the
joint and kinetic chain levels. Unlike existing methods that implicitly model
interactions using full-body poses as tokens, we argue that explicitly modeling
joint-level interactions is more natural and effective for generating realistic
HOIs, as it directly captures the geometric and semantic relationships between
joints, rather than modeling interactions in the latent pose space. To this
end, ChainHOI introduces a novel joint graph to capture potential interactions
with objects, and a Generative Spatiotemporal Graph Convolution Network to
explicitly model interactions at the joint level. Furthermore, we propose a
Kinematics-based Interaction Module that explicitly models interactions at the
kinetic chain level, ensuring more realistic and biomechanically coherent
motions. Evaluations on two public datasets demonstrate that ChainHOI
significantly outperforms previous methods, generating more realistic, and
semantically consistent HOIs. Code is available
\href{https://github.com/qinghuannn/ChainHOI}{here}.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:55:34 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zeng",
"Ling-An",
""
],
[
"Huang",
"Guohong",
""
],
[
"Wei",
"Yi-Lin",
""
],
[
"Gu",
"Shengbo",
""
],
[
"Tang",
"Yu-Ming",
""
],
[
"Meng",
"Jingke",
""
],
[
"Zheng",
"Wei-Shi",
""
]
] | TITLE: ChainHOI: Joint-based Kinematic Chain Modeling for Human-Object
Interaction Generation
ABSTRACT: We propose ChainHOI, a novel approach for text-driven human-object
interaction (HOI) generation that explicitly models interactions at both the
joint and kinetic chain levels. Unlike existing methods that implicitly model
interactions using full-body poses as tokens, we argue that explicitly modeling
joint-level interactions is more natural and effective for generating realistic
HOIs, as it directly captures the geometric and semantic relationships between
joints, rather than modeling interactions in the latent pose space. To this
end, ChainHOI introduces a novel joint graph to capture potential interactions
with objects, and a Generative Spatiotemporal Graph Convolution Network to
explicitly model interactions at the joint level. Furthermore, we propose a
Kinematics-based Interaction Module that explicitly models interactions at the
kinetic chain level, ensuring more realistic and biomechanically coherent
motions. Evaluations on two public datasets demonstrate that ChainHOI
significantly outperforms previous methods, generating more realistic, and
semantically consistent HOIs. Code is available
\href{https://github.com/qinghuannn/ChainHOI}{here}.
|
2503.13134 | Prakhar Bhardwaj | Prakhar Bhardwaj, Sheethal Bhat, Andreas Maier | Enhancing zero-shot learning in medical imaging: integrating clip with
advanced techniques for improved chest x-ray analysis | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Due to the large volume of medical imaging data, advanced AI methodologies
are needed to assist radiologists in diagnosing thoracic diseases from chest
X-rays (CXRs). Existing deep learning models often require large, labeled
datasets, which are scarce in medical imaging due to the time-consuming and
expert-driven annotation process. In this paper, we extend the existing
approach to enhance zero-shot learning in medical imaging by integrating
Contrastive Language-Image Pre-training (CLIP) with Momentum Contrast (MoCo),
resulting in our proposed model, MoCoCLIP. Our method addresses challenges
posed by class-imbalanced and unlabeled datasets, enabling improved detection
of pulmonary pathologies. Experimental results on the NIH ChestXray14 dataset
demonstrate that MoCoCLIP outperforms the state-of-the-art CheXZero model,
achieving relative improvement of approximately 6.5%. Furthermore, on the
CheXpert dataset, MoCoCLIP demonstrates superior zero-shot performance,
achieving an average AUC of 0.750 compared to CheXZero with 0.746 AUC,
highlighting its enhanced generalization capabilities on unseen data.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:59:34 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Bhardwaj",
"Prakhar",
""
],
[
"Bhat",
"Sheethal",
""
],
[
"Maier",
"Andreas",
""
]
] | TITLE: Enhancing zero-shot learning in medical imaging: integrating clip with
advanced techniques for improved chest x-ray analysis
ABSTRACT: Due to the large volume of medical imaging data, advanced AI methodologies
are needed to assist radiologists in diagnosing thoracic diseases from chest
X-rays (CXRs). Existing deep learning models often require large, labeled
datasets, which are scarce in medical imaging due to the time-consuming and
expert-driven annotation process. In this paper, we extend the existing
approach to enhance zero-shot learning in medical imaging by integrating
Contrastive Language-Image Pre-training (CLIP) with Momentum Contrast (MoCo),
resulting in our proposed model, MoCoCLIP. Our method addresses challenges
posed by class-imbalanced and unlabeled datasets, enabling improved detection
of pulmonary pathologies. Experimental results on the NIH ChestXray14 dataset
demonstrate that MoCoCLIP outperforms the state-of-the-art CheXZero model,
achieving relative improvement of approximately 6.5%. Furthermore, on the
CheXpert dataset, MoCoCLIP demonstrates superior zero-shot performance,
achieving an average AUC of 0.750 compared to CheXZero with 0.746 AUC,
highlighting its enhanced generalization capabilities on unseen data.
|
2503.13156 | Youssef Mourchid | Zakariae Zrimek, Youssef Mourchid, Mohammed El Hassouni | DynSTG-Mamba: Dynamic Spatio-Temporal Graph Mamba with Cross-Graph
Knowledge Distillation for Gait Disorders Recognition | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Gait disorder recognition plays a crucial role in the early diagnosis and
monitoring of movement disorders. Existing approaches, including
spatio-temporal graph convolutional networks (ST-GCNs), often face high memory
demands and struggle to capture complex spatio-temporal dependencies, limiting
their efficiency in clinical applications. To address these challenges, we
introduce DynSTG-Mamba (Dynamic Spatio-Temporal Graph Mamba), a novel framework
that combines DF-STGNN and STG-Mamba to enhance motion sequence modeling. The
DF-STGNN incorporates a dynamic spatio-temporal filter that adaptively adjusts
spatial connections between skeletal joints and temporal interactions across
different movement phases. This approach ensures better feature propagation
through dynamic graph structures by considering the hierarchical nature and
dynamics of skeletal gait data. Meanwhile, STG-Mamba, an extension of Mamba
adapted for skeletal motion data, ensures a continuous propagation of states,
facilitating the capture of long-term dependencies while reducing computational
complexity. To reduce the number of model parameters and computational costs
while maintaining consistency, we propose Cross-Graph Relational Knowledge
Distillation, a novel knowledge transfer mechanism that aligns relational
information between teacher (large architecture) and student models (small
architecture) while using shared memory. This ensures that the interactions and
movement patterns of the joints are accurately preserved in the motion
sequences. We validate our DynSTG-Mamba on KOA-NM, PD-WALK, and ATAXIA
datasets, where it outperforms state-of-the-art approaches by achieving in
terms of Accuracy, F1-score, and Recall. Our results highlight the efficiency
and robustness of our approach, offering a lightweight yet highly accurate
solution for automated gait analysis and movement disorder assessment.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:26:47 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zrimek",
"Zakariae",
""
],
[
"Mourchid",
"Youssef",
""
],
[
"Hassouni",
"Mohammed El",
""
]
] | TITLE: DynSTG-Mamba: Dynamic Spatio-Temporal Graph Mamba with Cross-Graph
Knowledge Distillation for Gait Disorders Recognition
ABSTRACT: Gait disorder recognition plays a crucial role in the early diagnosis and
monitoring of movement disorders. Existing approaches, including
spatio-temporal graph convolutional networks (ST-GCNs), often face high memory
demands and struggle to capture complex spatio-temporal dependencies, limiting
their efficiency in clinical applications. To address these challenges, we
introduce DynSTG-Mamba (Dynamic Spatio-Temporal Graph Mamba), a novel framework
that combines DF-STGNN and STG-Mamba to enhance motion sequence modeling. The
DF-STGNN incorporates a dynamic spatio-temporal filter that adaptively adjusts
spatial connections between skeletal joints and temporal interactions across
different movement phases. This approach ensures better feature propagation
through dynamic graph structures by considering the hierarchical nature and
dynamics of skeletal gait data. Meanwhile, STG-Mamba, an extension of Mamba
adapted for skeletal motion data, ensures a continuous propagation of states,
facilitating the capture of long-term dependencies while reducing computational
complexity. To reduce the number of model parameters and computational costs
while maintaining consistency, we propose Cross-Graph Relational Knowledge
Distillation, a novel knowledge transfer mechanism that aligns relational
information between teacher (large architecture) and student models (small
architecture) while using shared memory. This ensures that the interactions and
movement patterns of the joints are accurately preserved in the motion
sequences. We validate our DynSTG-Mamba on KOA-NM, PD-WALK, and ATAXIA
datasets, where it outperforms state-of-the-art approaches by achieving in
terms of Accuracy, F1-score, and Recall. Our results highlight the efficiency
and robustness of our approach, offering a lightweight yet highly accurate
solution for automated gait analysis and movement disorder assessment.
|
2503.13158 | Bernd Zimmering | Bernd Zimmering, Cec\'ilia Coelho, Vaibhav Gupta, Maria Maleshkova,
Oliver Niggemann | Laplace-Net: Learning Dynamical Systems with External Forcing | Preprint - under review | null | null | null | cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modelling forced dynamical systems - where an external input drives the
system state - is critical across diverse domains such as engineering, finance,
and the natural sciences. In this work, we propose Laplace-Net, a decoupled,
solver-free neural framework for learning forced and delay-aware systems. It
leverages a Laplace transform-based approach to decompose internal dynamics,
external inputs, and initial values into established theoretical concepts,
enhancing interpretability. Laplace-Net promotes transferability since the
system can be rapidly re-trained or fine-tuned for new forcing signals,
providing flexibility in applications ranging from controller adaptation to
long-horizon forecasting. Experimental results on eight benchmark datasets -
including linear, non-linear, and delayed systems - demonstrate the method's
improved accuracy and robustness compared to state-of-the-art approaches,
particularly in handling complex and previously unseen inputs.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:31:12 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zimmering",
"Bernd",
""
],
[
"Coelho",
"Cecília",
""
],
[
"Gupta",
"Vaibhav",
""
],
[
"Maleshkova",
"Maria",
""
],
[
"Niggemann",
"Oliver",
""
]
] | TITLE: Laplace-Net: Learning Dynamical Systems with External Forcing
ABSTRACT: Modelling forced dynamical systems - where an external input drives the
system state - is critical across diverse domains such as engineering, finance,
and the natural sciences. In this work, we propose Laplace-Net, a decoupled,
solver-free neural framework for learning forced and delay-aware systems. It
leverages a Laplace transform-based approach to decompose internal dynamics,
external inputs, and initial values into established theoretical concepts,
enhancing interpretability. Laplace-Net promotes transferability since the
system can be rapidly re-trained or fine-tuned for new forcing signals,
providing flexibility in applications ranging from controller adaptation to
long-horizon forecasting. Experimental results on eight benchmark datasets -
including linear, non-linear, and delayed systems - demonstrate the method's
improved accuracy and robustness compared to state-of-the-art approaches,
particularly in handling complex and previously unseen inputs.
|
2503.13160 | Zihao Liu | Zihao Liu, Xiaoyu Wu, Jianqin Wu, Xuxu Wang, Linlin Yang | Language-guided Open-world Video Anomaly Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video anomaly detection models aim to detect anomalies that deviate from what
is expected. In open-world scenarios, the expected events may change as
requirements change. For example, not wearing a mask is considered abnormal
during a flu outbreak but normal otherwise. However, existing methods assume
that the definition of anomalies is invariable, and thus are not applicable to
the open world. To address this, we propose a novel open-world VAD paradigm
with variable definitions, allowing guided detection through user-provided
natural language at inference time. This paradigm necessitates establishing a
robust mapping from video and textual definition to anomaly score. Therefore,
we propose LaGoVAD (Language-guided Open-world VAD), a model that dynamically
adapts anomaly definitions through two regularization strategies: diversifying
the relative durations of anomalies via dynamic video synthesis, and enhancing
feature robustness through contrastive learning with negative mining. Training
such adaptable models requires diverse anomaly definitions, but existing
datasets typically provide given labels without semantic descriptions. To
bridge this gap, we collect PreVAD (Pre-training Video Anomaly Dataset), the
largest and most diverse video anomaly dataset to date, featuring 35,279
annotated videos with multi-level category labels and descriptions that
explicitly define anomalies. Zero-shot experiments on seven datasets
demonstrate SOTA performance. Data and code will be released.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:31:19 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Liu",
"Zihao",
""
],
[
"Wu",
"Xiaoyu",
""
],
[
"Wu",
"Jianqin",
""
],
[
"Wang",
"Xuxu",
""
],
[
"Yang",
"Linlin",
""
]
] | TITLE: Language-guided Open-world Video Anomaly Detection
ABSTRACT: Video anomaly detection models aim to detect anomalies that deviate from what
is expected. In open-world scenarios, the expected events may change as
requirements change. For example, not wearing a mask is considered abnormal
during a flu outbreak but normal otherwise. However, existing methods assume
that the definition of anomalies is invariable, and thus are not applicable to
the open world. To address this, we propose a novel open-world VAD paradigm
with variable definitions, allowing guided detection through user-provided
natural language at inference time. This paradigm necessitates establishing a
robust mapping from video and textual definition to anomaly score. Therefore,
we propose LaGoVAD (Language-guided Open-world VAD), a model that dynamically
adapts anomaly definitions through two regularization strategies: diversifying
the relative durations of anomalies via dynamic video synthesis, and enhancing
feature robustness through contrastive learning with negative mining. Training
such adaptable models requires diverse anomaly definitions, but existing
datasets typically provide given labels without semantic descriptions. To
bridge this gap, we collect PreVAD (Pre-training Video Anomaly Dataset), the
largest and most diverse video anomaly dataset to date, featuring 35,279
annotated videos with multi-level category labels and descriptions that
explicitly define anomalies. Zero-shot experiments on seven datasets
demonstrate SOTA performance. Data and code will be released.
|
2503.13163 | Shani Gamrian | Shani Gamrian, Hila Barel, Feiran Li, Masakazu Yoshimura, Daisuke Iso | Beyond RGB: Adaptive Parallel Processing for RAW Object Detection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Object detection models are typically applied to standard RGB images
processed through Image Signal Processing (ISP) pipelines, which are designed
to enhance sensor-captured RAW images for human vision. However, these ISP
functions can lead to a loss of critical information that may be essential in
optimizing for computer vision tasks, such as object detection. In this work,
we introduce Raw Adaptation Module (RAM), a module designed to replace the
traditional ISP, with parameters optimized specifically for RAW object
detection. Inspired by the parallel processing mechanisms of the human visual
system, RAM departs from existing learned ISP methods by applying multiple ISP
functions in parallel rather than sequentially, allowing for a more
comprehensive capture of image features. These processed representations are
then fused in a specialized module, which dynamically integrates and optimizes
the information for the target task. This novel approach not only leverages the
full potential of RAW sensor data but also enables task-specific
pre-processing, resulting in superior object detection performance. Our
approach outperforms RGB-based methods and achieves state-of-the-art results
across diverse RAW image datasets under varying lighting conditions and dynamic
ranges.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:36:49 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Gamrian",
"Shani",
""
],
[
"Barel",
"Hila",
""
],
[
"Li",
"Feiran",
""
],
[
"Yoshimura",
"Masakazu",
""
],
[
"Iso",
"Daisuke",
""
]
] | TITLE: Beyond RGB: Adaptive Parallel Processing for RAW Object Detection
ABSTRACT: Object detection models are typically applied to standard RGB images
processed through Image Signal Processing (ISP) pipelines, which are designed
to enhance sensor-captured RAW images for human vision. However, these ISP
functions can lead to a loss of critical information that may be essential in
optimizing for computer vision tasks, such as object detection. In this work,
we introduce Raw Adaptation Module (RAM), a module designed to replace the
traditional ISP, with parameters optimized specifically for RAW object
detection. Inspired by the parallel processing mechanisms of the human visual
system, RAM departs from existing learned ISP methods by applying multiple ISP
functions in parallel rather than sequentially, allowing for a more
comprehensive capture of image features. These processed representations are
then fused in a specialized module, which dynamically integrates and optimizes
the information for the target task. This novel approach not only leverages the
full potential of RAW sensor data but also enables task-specific
pre-processing, resulting in superior object detection performance. Our
approach outperforms RGB-based methods and achieves state-of-the-art results
across diverse RAW image datasets under varying lighting conditions and dynamic
ranges.
|
2503.13184 | Shihao Yuan | Yuanze Li, Shihao Yuan, Haolin Wang, Qizhang Li, Ming Liu, Chen Xu,
Guangming Shi, Wangmeng Zuo | Triad: Empowering LMM-based Anomaly Detection with Vision Expert-guided
Visual Tokenizer and Manufacturing Process | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although recent methods have tried to introduce large multimodal models
(LMMs) into industrial anomaly detection (IAD), their generalization in the IAD
field is far inferior to that for general purposes. We summarize the main
reasons for this gap into two aspects. On one hand, general-purpose LMMs lack
cognition of defects in the visual modality, thereby failing to sufficiently
focus on defect areas. Therefore, we propose to modify the AnyRes structure of
the LLaVA model, providing the potential anomalous areas identified by existing
IAD models to the LMMs. On the other hand, existing methods mainly focus on
identifying defects by learning defect patterns or comparing with normal
samples, yet they fall short of understanding the causes of these defects.
Considering that the generation of defects is closely related to the
manufacturing process, we propose a manufacturing-driven IAD paradigm. An
instruction-tuning dataset for IAD (InstructIAD) and a data organization
approach for Chain-of-Thought with manufacturing (CoT-M) are designed to
leverage the manufacturing process for IAD. Based on the above two
modifications, we present Triad, a novel LMM-based method incorporating an
expert-guided region-of-interest tokenizer and manufacturing process for
industrial anomaly detection. Extensive experiments show that our Triad not
only demonstrates competitive performance against current LMMs but also
achieves further improved accuracy when equipped with manufacturing processes.
Source code, training data, and pre-trained models will be publicly available
at https://github.com/tzjtatata/Triad.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:56:57 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Li",
"Yuanze",
""
],
[
"Yuan",
"Shihao",
""
],
[
"Wang",
"Haolin",
""
],
[
"Li",
"Qizhang",
""
],
[
"Liu",
"Ming",
""
],
[
"Xu",
"Chen",
""
],
[
"Shi",
"Guangming",
""
],
[
"Zuo",
"Wangmeng",
""
]
] | TITLE: Triad: Empowering LMM-based Anomaly Detection with Vision Expert-guided
Visual Tokenizer and Manufacturing Process
ABSTRACT: Although recent methods have tried to introduce large multimodal models
(LMMs) into industrial anomaly detection (IAD), their generalization in the IAD
field is far inferior to that for general purposes. We summarize the main
reasons for this gap into two aspects. On one hand, general-purpose LMMs lack
cognition of defects in the visual modality, thereby failing to sufficiently
focus on defect areas. Therefore, we propose to modify the AnyRes structure of
the LLaVA model, providing the potential anomalous areas identified by existing
IAD models to the LMMs. On the other hand, existing methods mainly focus on
identifying defects by learning defect patterns or comparing with normal
samples, yet they fall short of understanding the causes of these defects.
Considering that the generation of defects is closely related to the
manufacturing process, we propose a manufacturing-driven IAD paradigm. An
instruction-tuning dataset for IAD (InstructIAD) and a data organization
approach for Chain-of-Thought with manufacturing (CoT-M) are designed to
leverage the manufacturing process for IAD. Based on the above two
modifications, we present Triad, a novel LMM-based method incorporating an
expert-guided region-of-interest tokenizer and manufacturing process for
industrial anomaly detection. Extensive experiments show that our Triad not
only demonstrates competitive performance against current LMMs but also
achieves further improved accuracy when equipped with manufacturing processes.
Source code, training data, and pre-trained models will be publicly available
at https://github.com/tzjtatata/Triad.
|
2503.13185 | Cheng Wang | Dingning Liu, Cheng Wang, Peng Gao, Renrui Zhang, Xinzhu Ma, Yuan
Meng, Zhihui Wang | 3DAxisPrompt: Promoting the 3D Grounding and Reasoning in GPT-4o | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Multimodal Large Language Models (MLLMs) exhibit impressive capabilities
across a variety of tasks, especially when equipped with carefully designed
visual prompts. However, existing studies primarily focus on logical reasoning
and visual understanding, while the capability of MLLMs to operate effectively
in 3D vision remains an ongoing area of exploration. In this paper, we
introduce a novel visual prompting method, called 3DAxisPrompt, to elicit the
3D understanding capabilities of MLLMs in real-world scenes. More specifically,
our method leverages the 3D coordinate axis and masks generated from the
Segment Anything Model (SAM) to provide explicit geometric priors to MLLMs and
then extend their impressive 2D grounding and reasoning ability to real-world
3D scenarios. Besides, we first provide a thorough investigation of the
potential visual prompting formats and conclude our findings to reveal the
potential and limits of 3D understanding capabilities in GPT-4o, as a
representative of MLLMs. Finally, we build evaluation environments with four
datasets, i.e., ScanRefer, ScanNet, FMB, and nuScene datasets, covering various
3D tasks. Based on this, we conduct extensive quantitative and qualitative
experiments, which demonstrate the effectiveness of the proposed method.
Overall, our study reveals that MLLMs, with the help of 3DAxisPrompt, can
effectively perceive an object's 3D position in real-world scenarios.
Nevertheless, a single prompt engineering approach does not consistently
achieve the best outcomes for all 3D tasks. This study highlights the
feasibility of leveraging MLLMs for 3D vision grounding/reasoning with prompt
engineering techniques.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:57:05 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Liu",
"Dingning",
""
],
[
"Wang",
"Cheng",
""
],
[
"Gao",
"Peng",
""
],
[
"Zhang",
"Renrui",
""
],
[
"Ma",
"Xinzhu",
""
],
[
"Meng",
"Yuan",
""
],
[
"Wang",
"Zhihui",
""
]
] | TITLE: 3DAxisPrompt: Promoting the 3D Grounding and Reasoning in GPT-4o
ABSTRACT: Multimodal Large Language Models (MLLMs) exhibit impressive capabilities
across a variety of tasks, especially when equipped with carefully designed
visual prompts. However, existing studies primarily focus on logical reasoning
and visual understanding, while the capability of MLLMs to operate effectively
in 3D vision remains an ongoing area of exploration. In this paper, we
introduce a novel visual prompting method, called 3DAxisPrompt, to elicit the
3D understanding capabilities of MLLMs in real-world scenes. More specifically,
our method leverages the 3D coordinate axis and masks generated from the
Segment Anything Model (SAM) to provide explicit geometric priors to MLLMs and
then extend their impressive 2D grounding and reasoning ability to real-world
3D scenarios. Besides, we first provide a thorough investigation of the
potential visual prompting formats and conclude our findings to reveal the
potential and limits of 3D understanding capabilities in GPT-4o, as a
representative of MLLMs. Finally, we build evaluation environments with four
datasets, i.e., ScanRefer, ScanNet, FMB, and nuScene datasets, covering various
3D tasks. Based on this, we conduct extensive quantitative and qualitative
experiments, which demonstrate the effectiveness of the proposed method.
Overall, our study reveals that MLLMs, with the help of 3DAxisPrompt, can
effectively perceive an object's 3D position in real-world scenarios.
Nevertheless, a single prompt engineering approach does not consistently
achieve the best outcomes for all 3D tasks. This study highlights the
feasibility of leveraging MLLMs for 3D vision grounding/reasoning with prompt
engineering techniques.
|
2503.13188 | Matteo Sodano | Matteo Sodano, Federico Magistri, Elias Marks, Fares Hosn, Aibek
Zurbayev, Rodrigo Marcuzzi, Meher V. R. Malladi, Jens Behley, Cyrill
Stachniss | 3D Hierarchical Panoptic Segmentation in Real Orchard Environments
Across Different Sensors | Submitted to IROS | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Crop yield estimation is a relevant problem in agriculture, because an
accurate crop yield estimate can support farmers' decisions on harvesting or
precision intervention. Robots can help to automate this process. To do so,
they need to be able to perceive the surrounding environment to identify target
objects. In this paper, we introduce a novel approach to address the problem of
hierarchical panoptic segmentation of apple orchards on 3D data from different
sensors. Our approach is able to simultaneously provide semantic segmentation,
instance segmentation of trunks and fruits, and instance segmentation of plants
(a single trunk with its fruits). This allows us to identify relevant
information such as individual plants, fruits, and trunks, and capture the
relationship among them, such as precisely estimate the number of fruits
associated to each tree in an orchard. Additionally, to efficiently evaluate
our approach for hierarchical panoptic segmentation, we provide a dataset
designed specifically for this task. Our dataset is recorded in Bonn in a real
apple orchard with a variety of sensors, spanning from a terrestrial laser
scanner to a RGB-D camera mounted on different robotic platforms. The
experiments show that our approach surpasses state-of-the-art approaches in 3D
panoptic segmentation in the agricultural domain, while also providing full
hierarchical panoptic segmentation. Our dataset has been made publicly
available at https://www.ipb.uni-bonn.de/data/hops/. We will provide the
open-source implementation of our approach and public competiton for
hierarchical panoptic segmentation on the hidden test sets upon paper
acceptance.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:59:20 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Sodano",
"Matteo",
""
],
[
"Magistri",
"Federico",
""
],
[
"Marks",
"Elias",
""
],
[
"Hosn",
"Fares",
""
],
[
"Zurbayev",
"Aibek",
""
],
[
"Marcuzzi",
"Rodrigo",
""
],
[
"Malladi",
"Meher V. R.",
""
],
[
"Behley",
"Jens",
""
],
[
"Stachniss",
"Cyrill",
""
]
] | TITLE: 3D Hierarchical Panoptic Segmentation in Real Orchard Environments
Across Different Sensors
ABSTRACT: Crop yield estimation is a relevant problem in agriculture, because an
accurate crop yield estimate can support farmers' decisions on harvesting or
precision intervention. Robots can help to automate this process. To do so,
they need to be able to perceive the surrounding environment to identify target
objects. In this paper, we introduce a novel approach to address the problem of
hierarchical panoptic segmentation of apple orchards on 3D data from different
sensors. Our approach is able to simultaneously provide semantic segmentation,
instance segmentation of trunks and fruits, and instance segmentation of plants
(a single trunk with its fruits). This allows us to identify relevant
information such as individual plants, fruits, and trunks, and capture the
relationship among them, such as precisely estimate the number of fruits
associated to each tree in an orchard. Additionally, to efficiently evaluate
our approach for hierarchical panoptic segmentation, we provide a dataset
designed specifically for this task. Our dataset is recorded in Bonn in a real
apple orchard with a variety of sensors, spanning from a terrestrial laser
scanner to a RGB-D camera mounted on different robotic platforms. The
experiments show that our approach surpasses state-of-the-art approaches in 3D
panoptic segmentation in the agricultural domain, while also providing full
hierarchical panoptic segmentation. Our dataset has been made publicly
available at https://www.ipb.uni-bonn.de/data/hops/. We will provide the
open-source implementation of our approach and public competiton for
hierarchical panoptic segmentation on the hidden test sets upon paper
acceptance.
|
2503.13195 | Jianhua Pei | Haoqi Huang, Ping Wang, Jianhua Pei, Jiacheng Wang, Shahen Alexanian,
Dusit Niyato | Deep Learning Advancements in Anomaly Detection: A Comprehensive Survey | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid expansion of data from diverse sources has made anomaly detection
(AD) increasingly essential for identifying unexpected observations that may
signal system failures, security breaches, or fraud. As datasets become more
complex and high-dimensional, traditional detection methods struggle to
effectively capture intricate patterns. Advances in deep learning have made AD
methods more powerful and adaptable, improving their ability to handle
high-dimensional and unstructured data. This survey provides a comprehensive
review of over 180 recent studies, focusing on deep learning-based AD
techniques. We categorize and analyze these methods into reconstruction-based
and prediction-based approaches, highlighting their effectiveness in modeling
complex data distributions. Additionally, we explore the integration of
traditional and deep learning methods, highlighting how hybrid approaches
combine the interpretability of traditional techniques with the flexibility of
deep learning to enhance detection accuracy and model transparency. Finally, we
identify open issues and propose future research directions to advance the
field of AD. This review bridges gaps in existing literature and serves as a
valuable resource for researchers and practitioners seeking to enhance AD
techniques using deep learning.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:04:48 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Huang",
"Haoqi",
""
],
[
"Wang",
"Ping",
""
],
[
"Pei",
"Jianhua",
""
],
[
"Wang",
"Jiacheng",
""
],
[
"Alexanian",
"Shahen",
""
],
[
"Niyato",
"Dusit",
""
]
] | TITLE: Deep Learning Advancements in Anomaly Detection: A Comprehensive Survey
ABSTRACT: The rapid expansion of data from diverse sources has made anomaly detection
(AD) increasingly essential for identifying unexpected observations that may
signal system failures, security breaches, or fraud. As datasets become more
complex and high-dimensional, traditional detection methods struggle to
effectively capture intricate patterns. Advances in deep learning have made AD
methods more powerful and adaptable, improving their ability to handle
high-dimensional and unstructured data. This survey provides a comprehensive
review of over 180 recent studies, focusing on deep learning-based AD
techniques. We categorize and analyze these methods into reconstruction-based
and prediction-based approaches, highlighting their effectiveness in modeling
complex data distributions. Additionally, we explore the integration of
traditional and deep learning methods, highlighting how hybrid approaches
combine the interpretability of traditional techniques with the flexibility of
deep learning to enhance detection accuracy and model transparency. Finally, we
identify open issues and propose future research directions to advance the
field of AD. This review bridges gaps in existing literature and serves as a
valuable resource for researchers and practitioners seeking to enhance AD
techniques using deep learning.
|
2503.13203 | Corentin Sautier | Corentin Sautier, Gilles Puy, Alexandre Boulch, Renaud Marlet, Vincent
Lepetit | Clustering is back: Reaching state-of-the-art LiDAR instance
segmentation without training | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Panoptic segmentation of LiDAR point clouds is fundamental to outdoor scene
understanding, with autonomous driving being a primary application. While
state-of-the-art approaches typically rely on end-to-end deep learning
architectures and extensive manual annotations of instances, the significant
cost and time investment required for labeling large-scale point cloud datasets
remains a major bottleneck in this field. In this work, we demonstrate that
competitive panoptic segmentation can be achieved using only semantic labels,
with instances predicted without any training or annotations. Our method
achieves performance comparable to current state-of-the-art supervised methods
on standard benchmarks including SemanticKITTI and nuScenes, and outperforms
every publicly available method on SemanticKITTI as a drop-in instance head
replacement, while running in real-time on a single-threaded CPU and requiring
no instance labels. Our method is fully explainable, and requires no learning
or parameter tuning. Code is available at https://github.com/valeoai/Alpine/
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:12:08 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Sautier",
"Corentin",
""
],
[
"Puy",
"Gilles",
""
],
[
"Boulch",
"Alexandre",
""
],
[
"Marlet",
"Renaud",
""
],
[
"Lepetit",
"Vincent",
""
]
] | TITLE: Clustering is back: Reaching state-of-the-art LiDAR instance
segmentation without training
ABSTRACT: Panoptic segmentation of LiDAR point clouds is fundamental to outdoor scene
understanding, with autonomous driving being a primary application. While
state-of-the-art approaches typically rely on end-to-end deep learning
architectures and extensive manual annotations of instances, the significant
cost and time investment required for labeling large-scale point cloud datasets
remains a major bottleneck in this field. In this work, we demonstrate that
competitive panoptic segmentation can be achieved using only semantic labels,
with instances predicted without any training or annotations. Our method
achieves performance comparable to current state-of-the-art supervised methods
on standard benchmarks including SemanticKITTI and nuScenes, and outperforms
every publicly available method on SemanticKITTI as a drop-in instance head
replacement, while running in real-time on a single-threaded CPU and requiring
no instance labels. Our method is fully explainable, and requires no learning
or parameter tuning. Code is available at https://github.com/valeoai/Alpine/
|
2503.13205 | Zhen Chen | Zhen Chen, Zhihao Peng, Xusheng Liang, Cheng Wang, Peigan Liang,
Linsheng Zeng, Minjie Ju, Yixuan Yuan | MAP: Evaluation and Multi-Agent Enhancement of Large Language Models for
Inpatient Pathways | null | null | null | null | cs.AI cs.CL cs.CV cs.HC cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Inpatient pathways demand complex clinical decision-making based on
comprehensive patient information, posing critical challenges for clinicians.
Despite advancements in large language models (LLMs) in medical applications,
limited research focused on artificial intelligence (AI) inpatient pathways
systems, due to the lack of large-scale inpatient datasets. Moreover, existing
medical benchmarks typically concentrated on medical question-answering and
examinations, ignoring the multifaceted nature of clinical decision-making in
inpatient settings. To address these gaps, we first developed the Inpatient
Pathway Decision Support (IPDS) benchmark from the MIMIC-IV database,
encompassing 51,274 cases across nine triage departments and 17 major disease
categories alongside 16 standardized treatment options. Then, we proposed the
Multi-Agent Inpatient Pathways (MAP) framework to accomplish inpatient pathways
with three clinical agents, including a triage agent managing the patient
admission, a diagnosis agent serving as the primary decision maker at the
department, and a treatment agent providing treatment plans. Additionally, our
MAP framework includes a chief agent overseeing the inpatient pathways to guide
and promote these three clinician agents. Extensive experiments showed our MAP
improved the diagnosis accuracy by 25.10% compared to the state-of-the-art LLM
HuatuoGPT2-13B. It is worth noting that our MAP demonstrated significant
clinical compliance, outperforming three board-certified clinicians by 10%-12%,
establishing a foundation for inpatient pathways systems.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:14:28 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Chen",
"Zhen",
""
],
[
"Peng",
"Zhihao",
""
],
[
"Liang",
"Xusheng",
""
],
[
"Wang",
"Cheng",
""
],
[
"Liang",
"Peigan",
""
],
[
"Zeng",
"Linsheng",
""
],
[
"Ju",
"Minjie",
""
],
[
"Yuan",
"Yixuan",
""
]
] | TITLE: MAP: Evaluation and Multi-Agent Enhancement of Large Language Models for
Inpatient Pathways
ABSTRACT: Inpatient pathways demand complex clinical decision-making based on
comprehensive patient information, posing critical challenges for clinicians.
Despite advancements in large language models (LLMs) in medical applications,
limited research focused on artificial intelligence (AI) inpatient pathways
systems, due to the lack of large-scale inpatient datasets. Moreover, existing
medical benchmarks typically concentrated on medical question-answering and
examinations, ignoring the multifaceted nature of clinical decision-making in
inpatient settings. To address these gaps, we first developed the Inpatient
Pathway Decision Support (IPDS) benchmark from the MIMIC-IV database,
encompassing 51,274 cases across nine triage departments and 17 major disease
categories alongside 16 standardized treatment options. Then, we proposed the
Multi-Agent Inpatient Pathways (MAP) framework to accomplish inpatient pathways
with three clinical agents, including a triage agent managing the patient
admission, a diagnosis agent serving as the primary decision maker at the
department, and a treatment agent providing treatment plans. Additionally, our
MAP framework includes a chief agent overseeing the inpatient pathways to guide
and promote these three clinician agents. Extensive experiments showed our MAP
improved the diagnosis accuracy by 25.10% compared to the state-of-the-art LLM
HuatuoGPT2-13B. It is worth noting that our MAP demonstrated significant
clinical compliance, outperforming three board-certified clinicians by 10%-12%,
establishing a foundation for inpatient pathways systems.
|
2503.13211 | Marvin Seyfarth | Marvin Seyfarth, Salman Ul Hassan Dar, Isabelle Ayx, Matthias
Alexander Fink, Stefan O. Schoenberg, Hans-Ulrich Kauczor, Sandy Engelhardt | MedLoRD: A Medical Low-Resource Diffusion Model for High-Resolution 3D
CT Image Synthesis | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Advancements in AI for medical imaging offer significant potential. However,
their applications are constrained by the limited availability of data and the
reluctance of medical centers to share it due to patient privacy concerns.
Generative models present a promising solution by creating synthetic data as a
substitute for real patient data. However, medical images are typically
high-dimensional, and current state-of-the-art methods are often impractical
for computational resource-constrained healthcare environments. These models
rely on data sub-sampling, raising doubts about their feasibility and
real-world applicability. Furthermore, many of these models are evaluated on
quantitative metrics that alone can be misleading in assessing the image
quality and clinical meaningfulness of the generated images. To address this,
we introduce MedLoRD, a generative diffusion model designed for computational
resource-constrained environments. MedLoRD is capable of generating
high-dimensional medical volumes with resolutions up to
512$\times$512$\times$256, utilizing GPUs with only 24GB VRAM, which are
commonly found in standard desktop workstations. MedLoRD is evaluated across
multiple modalities, including Coronary Computed Tomography Angiography and
Lung Computed Tomography datasets. Extensive evaluations through radiological
evaluation, relative regional volume analysis, adherence to conditional masks,
and downstream tasks show that MedLoRD generates high-fidelity images closely
adhering to segmentation mask conditions, surpassing the capabilities of
current state-of-the-art generative models for medical image synthesis in
computational resource-constrained environments.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:22:49 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Seyfarth",
"Marvin",
""
],
[
"Dar",
"Salman Ul Hassan",
""
],
[
"Ayx",
"Isabelle",
""
],
[
"Fink",
"Matthias Alexander",
""
],
[
"Schoenberg",
"Stefan O.",
""
],
[
"Kauczor",
"Hans-Ulrich",
""
],
[
"Engelhardt",
"Sandy",
""
]
] | TITLE: MedLoRD: A Medical Low-Resource Diffusion Model for High-Resolution 3D
CT Image Synthesis
ABSTRACT: Advancements in AI for medical imaging offer significant potential. However,
their applications are constrained by the limited availability of data and the
reluctance of medical centers to share it due to patient privacy concerns.
Generative models present a promising solution by creating synthetic data as a
substitute for real patient data. However, medical images are typically
high-dimensional, and current state-of-the-art methods are often impractical
for computational resource-constrained healthcare environments. These models
rely on data sub-sampling, raising doubts about their feasibility and
real-world applicability. Furthermore, many of these models are evaluated on
quantitative metrics that alone can be misleading in assessing the image
quality and clinical meaningfulness of the generated images. To address this,
we introduce MedLoRD, a generative diffusion model designed for computational
resource-constrained environments. MedLoRD is capable of generating
high-dimensional medical volumes with resolutions up to
512$\times$512$\times$256, utilizing GPUs with only 24GB VRAM, which are
commonly found in standard desktop workstations. MedLoRD is evaluated across
multiple modalities, including Coronary Computed Tomography Angiography and
Lung Computed Tomography datasets. Extensive evaluations through radiological
evaluation, relative regional volume analysis, adherence to conditional masks,
and downstream tasks show that MedLoRD generates high-fidelity images closely
adhering to segmentation mask conditions, surpassing the capabilities of
current state-of-the-art generative models for medical image synthesis in
computational resource-constrained environments.
|
2503.13226 | Konstantinos Nikoletos | Konstantinos Nikoletos, Vasilis Efthymiou, George Papadakis, Kostas
Stefanidis | Auto-Configuring Entity Resolution Pipelines | null | null | null | null | cs.DB | http://creativecommons.org/licenses/by/4.0/ | The same real-world entity (e.g., a movie, a restaurant, a person) may be
described in various ways on different datasets. Entity Resolution (ER) aims to
find such different descriptions of the same entity, this way improving data
quality and, therefore, data value. However, an ER pipeline typically involves
several steps (e.g., blocking, similarity estimation, clustering), with each
step requiring its own configurations and tuning. The choice of the best
configuration, among a vast number of possible combinations, is a
dataset-specific and labor-intensive task both for novice and expert users,
while it often requires some ground truth knowledge of real matches. In this
work, we examine ways of automatically configuring a state of-the-art
end-to-end ER pipeline based on pre-trained language models under two settings:
(i) When ground truth is available. In this case, sampling strategies that are
typically used for hyperparameter optimization can significantly restrict the
search of the configuration space. We experimentally compare their relative
effectiveness and time efficiency, applying them to ER pipelines for the first
time. (ii) When no ground truth is available. In this case, labelled data
extracted from other datasets with available ground truth can be used to train
a regression model that predicts the relative effectiveness of parameter
configurations. Experimenting with 11 ER benchmark datasets, we evaluate the
relative performance of existing techniques that address each problem, but have
not been applied to ER before.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:41:37 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Nikoletos",
"Konstantinos",
""
],
[
"Efthymiou",
"Vasilis",
""
],
[
"Papadakis",
"George",
""
],
[
"Stefanidis",
"Kostas",
""
]
] | TITLE: Auto-Configuring Entity Resolution Pipelines
ABSTRACT: The same real-world entity (e.g., a movie, a restaurant, a person) may be
described in various ways on different datasets. Entity Resolution (ER) aims to
find such different descriptions of the same entity, this way improving data
quality and, therefore, data value. However, an ER pipeline typically involves
several steps (e.g., blocking, similarity estimation, clustering), with each
step requiring its own configurations and tuning. The choice of the best
configuration, among a vast number of possible combinations, is a
dataset-specific and labor-intensive task both for novice and expert users,
while it often requires some ground truth knowledge of real matches. In this
work, we examine ways of automatically configuring a state of-the-art
end-to-end ER pipeline based on pre-trained language models under two settings:
(i) When ground truth is available. In this case, sampling strategies that are
typically used for hyperparameter optimization can significantly restrict the
search of the configuration space. We experimentally compare their relative
effectiveness and time efficiency, applying them to ER pipelines for the first
time. (ii) When no ground truth is available. In this case, labelled data
extracted from other datasets with available ground truth can be used to train
a regression model that predicts the relative effectiveness of parameter
configurations. Experimenting with 11 ER benchmark datasets, we evaluate the
relative performance of existing techniques that address each problem, but have
not been applied to ER before.
|
2503.13229 | Yongkang Cheng | Yongkang Cheng, Shaoli Huang | HoloGest: Decoupled Diffusion and Motion Priors for Generating
Holisticly Expressive Co-speech Gestures | Accepted by 3DV 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Animating virtual characters with holistic co-speech gestures is a
challenging but critical task. Previous systems have primarily focused on the
weak correlation between audio and gestures, leading to physically unnatural
outcomes that degrade the user experience. To address this problem, we
introduce HoleGest, a novel neural network framework based on decoupled
diffusion and motion priors for the automatic generation of high-quality,
expressive co-speech gestures. Our system leverages large-scale human motion
datasets to learn a robust prior with low audio dependency and high motion
reliance, enabling stable global motion and detailed finger movements. To
improve the generation efficiency of diffusion-based models, we integrate
implicit joint constraints with explicit geometric and conditional constraints,
capturing complex motion distributions between large strides. This integration
significantly enhances generation speed while maintaining high-quality motion.
Furthermore, we design a shared embedding space for gesture-transcription text
alignment, enabling the generation of semantically correct gesture actions.
Extensive experiments and user feedback demonstrate the effectiveness and
potential applications of our model, with our method achieving a level of
realism close to the ground truth, providing an immersive user experience. Our
code, model, and demo are are available at
https://cyk990422.github.io/HoloGest.github.io/.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:42:31 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Cheng",
"Yongkang",
""
],
[
"Huang",
"Shaoli",
""
]
] | TITLE: HoloGest: Decoupled Diffusion and Motion Priors for Generating
Holisticly Expressive Co-speech Gestures
ABSTRACT: Animating virtual characters with holistic co-speech gestures is a
challenging but critical task. Previous systems have primarily focused on the
weak correlation between audio and gestures, leading to physically unnatural
outcomes that degrade the user experience. To address this problem, we
introduce HoleGest, a novel neural network framework based on decoupled
diffusion and motion priors for the automatic generation of high-quality,
expressive co-speech gestures. Our system leverages large-scale human motion
datasets to learn a robust prior with low audio dependency and high motion
reliance, enabling stable global motion and detailed finger movements. To
improve the generation efficiency of diffusion-based models, we integrate
implicit joint constraints with explicit geometric and conditional constraints,
capturing complex motion distributions between large strides. This integration
significantly enhances generation speed while maintaining high-quality motion.
Furthermore, we design a shared embedding space for gesture-transcription text
alignment, enabling the generation of semantically correct gesture actions.
Extensive experiments and user feedback demonstrate the effectiveness and
potential applications of our model, with our method achieving a level of
realism close to the ground truth, providing an immersive user experience. Our
code, model, and demo are are available at
https://cyk990422.github.io/HoloGest.github.io/.
|
2503.13244 | Weiwei Su | Weiwei Su, Shigefumi Hata, Hiroshi Kori, Hiroya Nakao, and Ryota
Kobayashi | Pairwise vs Higher-order interactions: Can we identify the interaction
type in coupled oscillators from time series? | 16 pages, 6 figures | null | null | null | nlin.CD physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Rhythmic phenomena, which are ubiquitous in biological systems, are typically
modelled as systems of coupled limit cycle oscillators. Recently, there has
been an increased interest in understanding the impact of higher-order
interactions on the population dynamics of coupled oscillators. Meanwhile,
estimating a mathematical model from experimental data is a vital step in
understanding the dynamics of real-world complex systems. In coupled oscillator
systems, identifying the type of interaction (e.g. pairwise or three-body) of a
network is challenging, because different interactions can induce similar
dynamical states and bifurcations. In this study, we have developed a method
based on the adaptive LASSO (Least Absolute Shrinkage and Selection Operator)
to infer the interactions between the oscillators from time series data. The
proposed method can successfully classify the type of interaction and infer the
probabilities of the existence of pairwise and three-body couplings. Through
systematic analysis of synthetic datasets, we have demonstrated that our method
outperforms two baseline methods, LASSO and OLS (Ordinary Least Squares), in
accurately inferring the topology and strength of couplings between
oscillators. Finally, we demonstrate the effectiveness of the proposed method
by applying it to the synthetic data of 100 oscillators. These results imply
that the proposed method is promising for identifying interactions from
rhythmic activities in real-world systems.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:57:59 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Su",
"Weiwei",
""
],
[
"Hata",
"Shigefumi",
""
],
[
"Kori",
"Hiroshi",
""
],
[
"Nakao",
"Hiroya",
""
],
[
"Kobayashi",
"Ryota",
""
]
] | TITLE: Pairwise vs Higher-order interactions: Can we identify the interaction
type in coupled oscillators from time series?
ABSTRACT: Rhythmic phenomena, which are ubiquitous in biological systems, are typically
modelled as systems of coupled limit cycle oscillators. Recently, there has
been an increased interest in understanding the impact of higher-order
interactions on the population dynamics of coupled oscillators. Meanwhile,
estimating a mathematical model from experimental data is a vital step in
understanding the dynamics of real-world complex systems. In coupled oscillator
systems, identifying the type of interaction (e.g. pairwise or three-body) of a
network is challenging, because different interactions can induce similar
dynamical states and bifurcations. In this study, we have developed a method
based on the adaptive LASSO (Least Absolute Shrinkage and Selection Operator)
to infer the interactions between the oscillators from time series data. The
proposed method can successfully classify the type of interaction and infer the
probabilities of the existence of pairwise and three-body couplings. Through
systematic analysis of synthetic datasets, we have demonstrated that our method
outperforms two baseline methods, LASSO and OLS (Ordinary Least Squares), in
accurately inferring the topology and strength of couplings between
oscillators. Finally, we demonstrate the effectiveness of the proposed method
by applying it to the synthetic data of 100 oscillators. These results imply
that the proposed method is promising for identifying interactions from
rhythmic activities in real-world systems.
|
2503.13252 | Jingqi Jiang | Jingqi Jiang, Shida Xu, Kaicheng Zhang, Jiyuan Wei, Jingyang Wang and
Sen Wang | Digital Beamforming Enhanced Radar Odometry | null | null | null | null | cs.RO eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radar has become an essential sensor for autonomous navigation, especially in
challenging environments where camera and LiDAR sensors fail. 4D single-chip
millimeter-wave radar systems, in particular, have drawn increasing attention
thanks to their ability to provide spatial and Doppler information with low
hardware cost and power consumption. However, most single-chip radar systems
using traditional signal processing, such as Fast Fourier Transform, suffer
from limited spatial resolution in radar detection, significantly limiting the
performance of radar-based odometry and Simultaneous Localization and Mapping
(SLAM) systems. In this paper, we develop a novel radar signal processing
pipeline that integrates spatial domain beamforming techniques, and extend it
to 3D Direction of Arrival estimation. Experiments using public datasets are
conducted to evaluate and compare the performance of our proposed signal
processing pipeline against traditional methodologies. These tests specifically
focus on assessing structural precision across diverse scenes and measuring
odometry accuracy in different radar odometry systems. This research
demonstrates the feasibility of achieving more accurate radar odometry by
simply replacing the standard FFT-based processing with the proposed pipeline.
The codes are available at GitHub*.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:08:21 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Jiang",
"Jingqi",
""
],
[
"Xu",
"Shida",
""
],
[
"Zhang",
"Kaicheng",
""
],
[
"Wei",
"Jiyuan",
""
],
[
"Wang",
"Jingyang",
""
],
[
"Wang",
"Sen",
""
]
] | TITLE: Digital Beamforming Enhanced Radar Odometry
ABSTRACT: Radar has become an essential sensor for autonomous navigation, especially in
challenging environments where camera and LiDAR sensors fail. 4D single-chip
millimeter-wave radar systems, in particular, have drawn increasing attention
thanks to their ability to provide spatial and Doppler information with low
hardware cost and power consumption. However, most single-chip radar systems
using traditional signal processing, such as Fast Fourier Transform, suffer
from limited spatial resolution in radar detection, significantly limiting the
performance of radar-based odometry and Simultaneous Localization and Mapping
(SLAM) systems. In this paper, we develop a novel radar signal processing
pipeline that integrates spatial domain beamforming techniques, and extend it
to 3D Direction of Arrival estimation. Experiments using public datasets are
conducted to evaluate and compare the performance of our proposed signal
processing pipeline against traditional methodologies. These tests specifically
focus on assessing structural precision across diverse scenes and measuring
odometry accuracy in different radar odometry systems. This research
demonstrates the feasibility of achieving more accurate radar odometry by
simply replacing the standard FFT-based processing with the proposed pipeline.
The codes are available at GitHub*.
|
2503.13260 | Navve Wasserman | Amit Zalcher, Navve Wasserman, Roman Beliy, Oliver Heinimann, Michal
Irani | Don't Judge Before You CLIP: A Unified Approach for Perceptual Tasks | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visual perceptual tasks aim to predict human judgment of images (e.g.,
emotions invoked by images, image quality assessment). Unlike objective tasks
such as object/scene recognition, perceptual tasks rely on subjective human
assessments, making its data-labeling difficult. The scarcity of such
human-annotated data results in small datasets leading to poor generalization.
Typically, specialized models were designed for each perceptual task, tailored
to its unique characteristics and its own training dataset. We propose a
unified architectural framework for solving multiple different perceptual tasks
leveraging CLIP as a prior. Our approach is based on recent cognitive findings
which indicate that CLIP correlates well with human judgment. While CLIP was
explicitly trained to align images and text, it implicitly also learned human
inclinations. We attribute this to the inclusion of human-written image
captions in CLIP's training data, which contain not only factual image
descriptions, but inevitably also human sentiments and emotions. This makes
CLIP a particularly strong prior for perceptual tasks. Accordingly, we suggest
that minimal adaptation of CLIP suffices for solving a variety of perceptual
tasks. Our simple unified framework employs a lightweight adaptation to
fine-tune CLIP to each task, without requiring any task-specific architectural
changes. We evaluate our approach on three tasks: (i) Image Memorability
Prediction, (ii) No-reference Image Quality Assessment, and (iii) Visual
Emotion Analysis. Our model achieves state-of-the-art results on all three
tasks, while demonstrating improved generalization across different datasets.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:15:31 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zalcher",
"Amit",
""
],
[
"Wasserman",
"Navve",
""
],
[
"Beliy",
"Roman",
""
],
[
"Heinimann",
"Oliver",
""
],
[
"Irani",
"Michal",
""
]
] | TITLE: Don't Judge Before You CLIP: A Unified Approach for Perceptual Tasks
ABSTRACT: Visual perceptual tasks aim to predict human judgment of images (e.g.,
emotions invoked by images, image quality assessment). Unlike objective tasks
such as object/scene recognition, perceptual tasks rely on subjective human
assessments, making its data-labeling difficult. The scarcity of such
human-annotated data results in small datasets leading to poor generalization.
Typically, specialized models were designed for each perceptual task, tailored
to its unique characteristics and its own training dataset. We propose a
unified architectural framework for solving multiple different perceptual tasks
leveraging CLIP as a prior. Our approach is based on recent cognitive findings
which indicate that CLIP correlates well with human judgment. While CLIP was
explicitly trained to align images and text, it implicitly also learned human
inclinations. We attribute this to the inclusion of human-written image
captions in CLIP's training data, which contain not only factual image
descriptions, but inevitably also human sentiments and emotions. This makes
CLIP a particularly strong prior for perceptual tasks. Accordingly, we suggest
that minimal adaptation of CLIP suffices for solving a variety of perceptual
tasks. Our simple unified framework employs a lightweight adaptation to
fine-tune CLIP to each task, without requiring any task-specific architectural
changes. We evaluate our approach on three tasks: (i) Image Memorability
Prediction, (ii) No-reference Image Quality Assessment, and (iii) Visual
Emotion Analysis. Our model achieves state-of-the-art results on all three
tasks, while demonstrating improved generalization across different datasets.
|
2503.13271 | Chengen Wang | Chengen Wang, Murat Kantarcioglu | Graph Generative Models Evaluation with Masked Autoencoder | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, numerous graph generative models (GGMs) have been proposed.
However, evaluating these models remains a considerable challenge, primarily
due to the difficulty in extracting meaningful graph features that accurately
represent real-world graphs. The traditional evaluation techniques, which rely
on graph statistical properties like node degree distribution, clustering
coefficients, or Laplacian spectrum, overlook node features and lack
scalability. There are newly proposed deep learning-based methods employing
graph random neural networks or contrastive learning to extract graph features,
demonstrating superior performance compared to traditional statistical methods,
but their experimental results also demonstrate that these methods do not
always working well across different metrics. Although there are overlaps among
these metrics, they are generally not interchangeable, each evaluating
generative models from a different perspective. In this paper, we propose a
novel method that leverages graph masked autoencoders to effectively extract
graph features for GGM evaluations. We conduct extensive experiments on graphs
and empirically demonstrate that our method can be more reliable and effective
than previously proposed methods across a number of GGM evaluation metrics,
such as "Fr\'echet Distance (FD)" and "MMD Linear". However, no single method
stands out consistently across all metrics and datasets. Therefore, this study
also aims to raise awareness of the significance and challenges associated with
GGM evaluation techniques, especially in light of recent advances in generative
models.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:23:21 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Wang",
"Chengen",
""
],
[
"Kantarcioglu",
"Murat",
""
]
] | TITLE: Graph Generative Models Evaluation with Masked Autoencoder
ABSTRACT: In recent years, numerous graph generative models (GGMs) have been proposed.
However, evaluating these models remains a considerable challenge, primarily
due to the difficulty in extracting meaningful graph features that accurately
represent real-world graphs. The traditional evaluation techniques, which rely
on graph statistical properties like node degree distribution, clustering
coefficients, or Laplacian spectrum, overlook node features and lack
scalability. There are newly proposed deep learning-based methods employing
graph random neural networks or contrastive learning to extract graph features,
demonstrating superior performance compared to traditional statistical methods,
but their experimental results also demonstrate that these methods do not
always working well across different metrics. Although there are overlaps among
these metrics, they are generally not interchangeable, each evaluating
generative models from a different perspective. In this paper, we propose a
novel method that leverages graph masked autoencoders to effectively extract
graph features for GGM evaluations. We conduct extensive experiments on graphs
and empirically demonstrate that our method can be more reliable and effective
than previously proposed methods across a number of GGM evaluation metrics,
such as "Fr\'echet Distance (FD)" and "MMD Linear". However, no single method
stands out consistently across all metrics and datasets. Therefore, this study
also aims to raise awareness of the significance and challenges associated with
GGM evaluation techniques, especially in light of recent advances in generative
models.
|
2503.13272 | Katja Schwarz | Katja Schwarz and Norman Mueller and Peter Kontschieder | Generative Gaussian Splatting: Generating 3D Scenes with Video Diffusion
Priors | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Synthesizing consistent and photorealistic 3D scenes is an open problem in
computer vision. Video diffusion models generate impressive videos but cannot
directly synthesize 3D representations, i.e., lack 3D consistency in the
generated sequences. In addition, directly training generative 3D models is
challenging due to a lack of 3D training data at scale. In this work, we
present Generative Gaussian Splatting (GGS) -- a novel approach that integrates
a 3D representation with a pre-trained latent video diffusion model.
Specifically, our model synthesizes a feature field parameterized via 3D
Gaussian primitives. The feature field is then either rendered to feature maps
and decoded into multi-view images, or directly upsampled into a 3D radiance
field. We evaluate our approach on two common benchmark datasets for scene
synthesis, RealEstate10K and ScanNet+, and find that our proposed GGS model
significantly improves both the 3D consistency of the generated multi-view
images, and the quality of the generated 3D scenes over all relevant baselines.
Compared to a similar model without 3D representation, GGS improves FID on the
generated 3D scenes by ~20% on both RealEstate10K and ScanNet+. Project page:
https://katjaschwarz.github.io/ggs/
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:24:04 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Schwarz",
"Katja",
""
],
[
"Mueller",
"Norman",
""
],
[
"Kontschieder",
"Peter",
""
]
] | TITLE: Generative Gaussian Splatting: Generating 3D Scenes with Video Diffusion
Priors
ABSTRACT: Synthesizing consistent and photorealistic 3D scenes is an open problem in
computer vision. Video diffusion models generate impressive videos but cannot
directly synthesize 3D representations, i.e., lack 3D consistency in the
generated sequences. In addition, directly training generative 3D models is
challenging due to a lack of 3D training data at scale. In this work, we
present Generative Gaussian Splatting (GGS) -- a novel approach that integrates
a 3D representation with a pre-trained latent video diffusion model.
Specifically, our model synthesizes a feature field parameterized via 3D
Gaussian primitives. The feature field is then either rendered to feature maps
and decoded into multi-view images, or directly upsampled into a 3D radiance
field. We evaluate our approach on two common benchmark datasets for scene
synthesis, RealEstate10K and ScanNet+, and find that our proposed GGS model
significantly improves both the 3D consistency of the generated multi-view
images, and the quality of the generated 3D scenes over all relevant baselines.
Compared to a similar model without 3D representation, GGS improves FID on the
generated 3D scenes by ~20% on both RealEstate10K and ScanNet+. Project page:
https://katjaschwarz.github.io/ggs/
|
2503.13277 | Suresh Kumar | Alfred Simbun, Suresh Kumar | Artificial Intelligence-Driven Prognostic Classification of COVID-19
Using Chest X-rays: A Deep Learning Approach | 27 pages, 6 figures, 10 tables | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Background: The COVID-19 pandemic has overwhelmed healthcare systems,
emphasizing the need for AI-driven tools to assist in rapid and accurate
patient prognosis. Chest X-ray imaging is a widely available diagnostic tool,
but existing methods for prognosis classification lack scalability and
efficiency. Objective: This study presents a high-accuracy deep learning model
for classifying COVID-19 severity (Mild, Moderate, and Severe) using Chest
X-ray images, developed on Microsoft Azure Custom Vision. Methods: Using a
dataset of 1,103 confirmed COVID-19 X-ray images from AIforCOVID, we trained
and validated a deep learning model leveraging Convolutional Neural Networks
(CNNs). The model was evaluated on an unseen dataset to measure accuracy,
precision, and recall. Results: Our model achieved an average accuracy of 97%,
with specificity of 99%, sensitivity of 87%, and an F1-score of 93.11%. When
classifying COVID-19 severity, the model achieved accuracies of 89.03% (Mild),
95.77% (Moderate), and 81.16% (Severe). These results demonstrate the model's
potential for real-world clinical applications, aiding in faster
decision-making and improved resource allocation. Conclusion: AI-driven
prognosis classification using deep learning can significantly enhance COVID-19
patient management, enabling early intervention and efficient triaging. Our
study provides a scalable, high-accuracy AI framework for integrating deep
learning into routine clinical workflows. Future work should focus on expanding
datasets, external validation, and regulatory compliance to facilitate clinical
adoption.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:27:21 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Simbun",
"Alfred",
""
],
[
"Kumar",
"Suresh",
""
]
] | TITLE: Artificial Intelligence-Driven Prognostic Classification of COVID-19
Using Chest X-rays: A Deep Learning Approach
ABSTRACT: Background: The COVID-19 pandemic has overwhelmed healthcare systems,
emphasizing the need for AI-driven tools to assist in rapid and accurate
patient prognosis. Chest X-ray imaging is a widely available diagnostic tool,
but existing methods for prognosis classification lack scalability and
efficiency. Objective: This study presents a high-accuracy deep learning model
for classifying COVID-19 severity (Mild, Moderate, and Severe) using Chest
X-ray images, developed on Microsoft Azure Custom Vision. Methods: Using a
dataset of 1,103 confirmed COVID-19 X-ray images from AIforCOVID, we trained
and validated a deep learning model leveraging Convolutional Neural Networks
(CNNs). The model was evaluated on an unseen dataset to measure accuracy,
precision, and recall. Results: Our model achieved an average accuracy of 97%,
with specificity of 99%, sensitivity of 87%, and an F1-score of 93.11%. When
classifying COVID-19 severity, the model achieved accuracies of 89.03% (Mild),
95.77% (Moderate), and 81.16% (Severe). These results demonstrate the model's
potential for real-world clinical applications, aiding in faster
decision-making and improved resource allocation. Conclusion: AI-driven
prognosis classification using deep learning can significantly enhance COVID-19
patient management, enabling early intervention and efficient triaging. Our
study provides a scalable, high-accuracy AI framework for integrating deep
learning into routine clinical workflows. Future work should focus on expanding
datasets, external validation, and regulatory compliance to facilitate clinical
adoption.
|
2503.13279 | Xinkai Zou | Xinkai Zou, Yan Liu, Xiongbo Shi, Chen Yang | Goal2Story: A Multi-Agent Fleet based on Privately Enabled sLLMs for
Impacting Mapping on Requirements Elicitation | null | null | null | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As requirements drift with rapid iterations, agile development becomes the
dominant paradigm. Goal-driven Requirements Elicitation (RE) is a pivotal yet
challenging task in agile project development due to its heavy tangling with
adaptive planning and efficient collaboration. Recently, AI agents have shown
promising ability in supporting requirements analysis by saving significant
time and effort for stakeholders. However, current research mainly focuses on
functional RE, and research works have not been reported bridging the long
journey from goal to user stories. Moreover, considering the cost of LLM
facilities and the need for data and idea protection, privately hosted
small-sized LLM should be further utilized in RE. To address these challenges,
we propose Goal2Story, a multi-agent fleet that adopts the Impact Mapping (IM)
framework while merely using cost-effective sLLMs for goal-driven RE. Moreover,
we introduce a StorySeek dataset that contains over 1,000 user stories (USs)
with corresponding goals and project context information, as well as the
semi-automatic dataset construction method. For evaluation, we proposed two
metrics: Factuality Hit Rate (FHR) to measure consistency between the generated
USs with the dataset and Quality And Consistency Evaluation (QuACE) to evaluate
the quality of the generated USs. Experimental results demonstrate that
Goal2Story outperforms the baseline performance of the Super-Agent adopting
powerful LLMs, while also showcasing the performance improvements in key
metrics brought by CoT and Agent Profile to Goal2Story, as well as its
exploration in identifying latent needs.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:31:20 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zou",
"Xinkai",
""
],
[
"Liu",
"Yan",
""
],
[
"Shi",
"Xiongbo",
""
],
[
"Yang",
"Chen",
""
]
] | TITLE: Goal2Story: A Multi-Agent Fleet based on Privately Enabled sLLMs for
Impacting Mapping on Requirements Elicitation
ABSTRACT: As requirements drift with rapid iterations, agile development becomes the
dominant paradigm. Goal-driven Requirements Elicitation (RE) is a pivotal yet
challenging task in agile project development due to its heavy tangling with
adaptive planning and efficient collaboration. Recently, AI agents have shown
promising ability in supporting requirements analysis by saving significant
time and effort for stakeholders. However, current research mainly focuses on
functional RE, and research works have not been reported bridging the long
journey from goal to user stories. Moreover, considering the cost of LLM
facilities and the need for data and idea protection, privately hosted
small-sized LLM should be further utilized in RE. To address these challenges,
we propose Goal2Story, a multi-agent fleet that adopts the Impact Mapping (IM)
framework while merely using cost-effective sLLMs for goal-driven RE. Moreover,
we introduce a StorySeek dataset that contains over 1,000 user stories (USs)
with corresponding goals and project context information, as well as the
semi-automatic dataset construction method. For evaluation, we proposed two
metrics: Factuality Hit Rate (FHR) to measure consistency between the generated
USs with the dataset and Quality And Consistency Evaluation (QuACE) to evaluate
the quality of the generated USs. Experimental results demonstrate that
Goal2Story outperforms the baseline performance of the Super-Agent adopting
powerful LLMs, while also showcasing the performance improvements in key
metrics brought by CoT and Agent Profile to Goal2Story, as well as its
exploration in identifying latent needs.
|
2503.13296 | Mikkel Jordahn | Mikkel Jordahn, Jonas Vestergaard Jensen, Mikkel N. Schmidt and
Michael Riis Andersen | On Local Posterior Structure in Deep Ensembles | Code and models available at
https://github.com/jonasvj/OnLocalPosteriorStructureInDeepEnsembles | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by-sa/4.0/ | Bayesian Neural Networks (BNNs) often improve model calibration and
predictive uncertainty quantification compared to point estimators such as
maximum-a-posteriori (MAP). Similarly, deep ensembles (DEs) are also known to
improve calibration, and therefore, it is natural to hypothesize that deep
ensembles of BNNs (DE-BNNs) should provide even further improvements. In this
work, we systematically investigate this across a number of datasets, neural
network architectures, and BNN approximation methods and surprisingly find that
when the ensembles grow large enough, DEs consistently outperform DE-BNNs on
in-distribution data. To shine light on this observation, we conduct several
sensitivity and ablation studies. Moreover, we show that even though DE-BNNs
outperform DEs on out-of-distribution metrics, this comes at the cost of
decreased in-distribution performance. As a final contribution, we open-source
the large pool of trained models to facilitate further research on this topic.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:41:39 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Jordahn",
"Mikkel",
""
],
[
"Jensen",
"Jonas Vestergaard",
""
],
[
"Schmidt",
"Mikkel N.",
""
],
[
"Andersen",
"Michael Riis",
""
]
] | TITLE: On Local Posterior Structure in Deep Ensembles
ABSTRACT: Bayesian Neural Networks (BNNs) often improve model calibration and
predictive uncertainty quantification compared to point estimators such as
maximum-a-posteriori (MAP). Similarly, deep ensembles (DEs) are also known to
improve calibration, and therefore, it is natural to hypothesize that deep
ensembles of BNNs (DE-BNNs) should provide even further improvements. In this
work, we systematically investigate this across a number of datasets, neural
network architectures, and BNN approximation methods and surprisingly find that
when the ensembles grow large enough, DEs consistently outperform DE-BNNs on
in-distribution data. To shine light on this observation, we conduct several
sensitivity and ablation studies. Moreover, we show that even though DE-BNNs
outperform DEs on out-of-distribution metrics, this comes at the cost of
decreased in-distribution performance. As a final contribution, we open-source
the large pool of trained models to facilitate further research on this topic.
|
2503.13301 | Deepak Vungarala | Deepak Vungarala, Md Hasibul Amin, Pietro Mercati, Arnob Ghosh, Arman
Roohi, Ramtin Zand, Shaahin Angizi | LIMCA: LLM for Automating Analog In-Memory Computing Architecture Design
Exploration | 4 Figures, 5 Tables | null | null | null | cs.AR | http://creativecommons.org/licenses/by/4.0/ | Resistive crossbars enabling analog In-Memory Computing (IMC) have emerged as
a promising architecture for Deep Neural Network (DNN) acceleration, offering
high memory bandwidth and in-situ computation. However, the manual,
knowledge-intensive design process and the lack of high-quality circuit
netlists have significantly constrained design space exploration and
optimization to behavioral system-level tools. In this work, we introduce
LIMCA, a novel fine-tune-free Large Language Model (LLM)-driven framework for
automating the design and evaluation of IMC crossbar architectures. Unlike
traditional approaches, LIMCA employs a No-Human-In-Loop (NHIL) automated
pipeline to generate and validate circuit netlists for SPICE simulations,
eliminating manual intervention. LIMCA systematically explores the IMC design
space by leveraging a structured dataset and LLM-based performance evaluation.
Our experimental results on MNIST classification demonstrate that LIMCA
successfully generates crossbar designs achieving $\geq$96% accuracy while
maintaining a power consumption $\leq$3W, making this the first work in
LLM-assisted IMC design space exploration. Compared to existing frameworks,
LIMCA provides an automated, scalable, and hardware-aware solution, reducing
design exploration time while ensuring user-constrained performance trade-offs.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:45:17 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Vungarala",
"Deepak",
""
],
[
"Amin",
"Md Hasibul",
""
],
[
"Mercati",
"Pietro",
""
],
[
"Ghosh",
"Arnob",
""
],
[
"Roohi",
"Arman",
""
],
[
"Zand",
"Ramtin",
""
],
[
"Angizi",
"Shaahin",
""
]
] | TITLE: LIMCA: LLM for Automating Analog In-Memory Computing Architecture Design
Exploration
ABSTRACT: Resistive crossbars enabling analog In-Memory Computing (IMC) have emerged as
a promising architecture for Deep Neural Network (DNN) acceleration, offering
high memory bandwidth and in-situ computation. However, the manual,
knowledge-intensive design process and the lack of high-quality circuit
netlists have significantly constrained design space exploration and
optimization to behavioral system-level tools. In this work, we introduce
LIMCA, a novel fine-tune-free Large Language Model (LLM)-driven framework for
automating the design and evaluation of IMC crossbar architectures. Unlike
traditional approaches, LIMCA employs a No-Human-In-Loop (NHIL) automated
pipeline to generate and validate circuit netlists for SPICE simulations,
eliminating manual intervention. LIMCA systematically explores the IMC design
space by leveraging a structured dataset and LLM-based performance evaluation.
Our experimental results on MNIST classification demonstrate that LIMCA
successfully generates crossbar designs achieving $\geq$96% accuracy while
maintaining a power consumption $\leq$3W, making this the first work in
LLM-assisted IMC design space exploration. Compared to existing frameworks,
LIMCA provides an automated, scalable, and hardware-aware solution, reducing
design exploration time while ensuring user-constrained performance trade-offs.
|
2503.13304 | Witold Wydma\'nski | Witold Wydma\'nski, Marek \'Smieja | GFSNetwork: Differentiable Feature Selection via Gumbel-Sigmoid
Relaxation | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Feature selection in deep learning remains a critical challenge, particularly
for high-dimensional tabular data where interpretability and computational
efficiency are paramount. We present GFSNetwork, a novel neural architecture
that performs differentiable feature selection through temperature-controlled
Gumbel-Sigmoid sampling. Unlike traditional methods, where the user has to
define the requested number of features, GFSNetwork selects it automatically
during an end-to-end process. Moreover, GFSNetwork maintains constant
computational overhead regardless of the number of input features. We evaluate
GFSNetwork on a series of classification and regression benchmarks, where it
consistently outperforms recent methods including DeepLasso, attention maps, as
well as traditional feature selectors, while using significantly fewer
features. Furthermore, we validate our approach on real-world metagenomic
datasets, demonstrating its effectiveness in high-dimensional biological data.
Concluding, our method provides a scalable solution that bridges the gap
between neural network flexibility and traditional feature selection
interpretability. We share our python implementation of GFSNetwork at
https://github.com/wwydmanski/GFSNetwork, as well as a PyPi package
(gfs_network).
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:47:26 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Wydmański",
"Witold",
""
],
[
"Śmieja",
"Marek",
""
]
] | TITLE: GFSNetwork: Differentiable Feature Selection via Gumbel-Sigmoid
Relaxation
ABSTRACT: Feature selection in deep learning remains a critical challenge, particularly
for high-dimensional tabular data where interpretability and computational
efficiency are paramount. We present GFSNetwork, a novel neural architecture
that performs differentiable feature selection through temperature-controlled
Gumbel-Sigmoid sampling. Unlike traditional methods, where the user has to
define the requested number of features, GFSNetwork selects it automatically
during an end-to-end process. Moreover, GFSNetwork maintains constant
computational overhead regardless of the number of input features. We evaluate
GFSNetwork on a series of classification and regression benchmarks, where it
consistently outperforms recent methods including DeepLasso, attention maps, as
well as traditional feature selectors, while using significantly fewer
features. Furthermore, we validate our approach on real-world metagenomic
datasets, demonstrating its effectiveness in high-dimensional biological data.
Concluding, our method provides a scalable solution that bridges the gap
between neural network flexibility and traditional feature selection
interpretability. We share our python implementation of GFSNetwork at
https://github.com/wwydmanski/GFSNetwork, as well as a PyPi package
(gfs_network).
|
2503.13310 | Matteo Esposito | Matteo Esposito and Xiaozhou Li and Sergio Moreschini and Noman Ahmad
and Tomas Cerny and Karthik Vaidhyanathan and Valentina Lenarduzzi and Davide
Taibi | Generative AI for Software Architecture. Applications, Trends,
Challenges, and Future Directions | null | null | null | null | cs.SE cs.AI cs.DC cs.ET | http://creativecommons.org/licenses/by/4.0/ | Context: Generative Artificial Intelligence (GenAI) is transforming much of
software development, yet its application in software architecture is still in
its infancy, and no prior study has systematically addressed the topic. Aim: We
aim to systematically synthesize the use, rationale, contexts, usability, and
future challenges of GenAI in software architecture. Method: We performed a
multivocal literature review (MLR), analyzing peer-reviewed and gray
literature, identifying current practices, models, adoption contexts, and
reported challenges, extracting themes via open coding. Results: Our review
identified significant adoption of GenAI for architectural decision support and
architectural reconstruction. OpenAI GPT models are predominantly applied, and
there is consistent use of techniques such as few-shot prompting and
retrieved-augmented generation (RAG). GenAI has been applied mostly to initial
stages of the Software Development Life Cycle (SDLC), such as
Requirements-to-Architecture and Architecture-to-Code. Monolithic and
microservice architectures were the dominant targets. However, rigorous testing
of GenAI outputs was typically missing from the studies. Among the most
frequent challenges are model precision, hallucinations, ethical aspects,
privacy issues, lack of architecture-specific datasets, and the absence of
sound evaluation frameworks. Conclusions: GenAI shows significant potential in
software design, but several challenges remain on its path to greater adoption.
Research efforts should target designing general evaluation methodologies,
handling ethics and precision, increasing transparency and explainability, and
promoting architecture-specific datasets and benchmarks to bridge the gap
between theoretical possibilities and practical use.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:49:30 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Esposito",
"Matteo",
""
],
[
"Li",
"Xiaozhou",
""
],
[
"Moreschini",
"Sergio",
""
],
[
"Ahmad",
"Noman",
""
],
[
"Cerny",
"Tomas",
""
],
[
"Vaidhyanathan",
"Karthik",
""
],
[
"Lenarduzzi",
"Valentina",
""
],
[
"Taibi",
"Davide",
""
]
] | TITLE: Generative AI for Software Architecture. Applications, Trends,
Challenges, and Future Directions
ABSTRACT: Context: Generative Artificial Intelligence (GenAI) is transforming much of
software development, yet its application in software architecture is still in
its infancy, and no prior study has systematically addressed the topic. Aim: We
aim to systematically synthesize the use, rationale, contexts, usability, and
future challenges of GenAI in software architecture. Method: We performed a
multivocal literature review (MLR), analyzing peer-reviewed and gray
literature, identifying current practices, models, adoption contexts, and
reported challenges, extracting themes via open coding. Results: Our review
identified significant adoption of GenAI for architectural decision support and
architectural reconstruction. OpenAI GPT models are predominantly applied, and
there is consistent use of techniques such as few-shot prompting and
retrieved-augmented generation (RAG). GenAI has been applied mostly to initial
stages of the Software Development Life Cycle (SDLC), such as
Requirements-to-Architecture and Architecture-to-Code. Monolithic and
microservice architectures were the dominant targets. However, rigorous testing
of GenAI outputs was typically missing from the studies. Among the most
frequent challenges are model precision, hallucinations, ethical aspects,
privacy issues, lack of architecture-specific datasets, and the absence of
sound evaluation frameworks. Conclusions: GenAI shows significant potential in
software design, but several challenges remain on its path to greater adoption.
Research efforts should target designing general evaluation methodologies,
handling ethics and precision, increasing transparency and explainability, and
promoting architecture-specific datasets and benchmarks to bridge the gap
between theoretical possibilities and practical use.
|
2503.13316 | Marcello Iotti | Marcello Iotti, Paolo Davini, Jost von Hardenberg, Giuseppe Zappa | RainScaleGAN: a Conditional Generative Adversarial Network for Rainfall
Downscaling | 38 pages, 16 figures | null | null | null | physics.ao-ph cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To this day, accurately simulating local-scale precipitation and reliably
reproducing its distribution remains a challenging task. The limited horizontal
resolution of Global Climate Models is among the primary factors undermining
their skill in this context. The physical mechanisms driving the onset and
development of precipitation, especially in extreme events, operate at
spatio-temporal scales smaller than those numerically resolved, thus struggling
to be captured accurately. In order to circumvent this limitation, several
downscaling approaches have been developed over the last decades to address the
discrepancy between the spatial resolution of models output and the resolution
required by local-scale applications. In this paper, we introduce RainScaleGAN,
a conditional deep convolutional Generative Adversarial Network (GAN) for
precipitation downscaling. GANs have been effectively used in image
super-resolution, an approach highly relevant for downscaling tasks.
RainScaleGAN's capabilities are tested in a perfect-model setup, where the
spatial resolution of a precipitation dataset is artificially degraded from
0.25$^{\circ}\times$0.25$^{\circ}$ to 2$^{\circ}\times$2$^\circ$, and
RainScaleGAN is used to restore it. The developed model outperforms one of the
leading precipitation downscaling method found in the literature. RainScaleGAN
not only generates a synthetic dataset featuring plausible high-resolution
spatial patterns and intensities, but also produces a precipitation
distribution with statistics closely mirroring those of the ground-truth
dataset. Given that RainScaleGAN's approach is agnostic with respect to the
underlying physics, the method has the potential to be applied to other
physical variables such as surface winds or temperature.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 15:54:20 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Iotti",
"Marcello",
""
],
[
"Davini",
"Paolo",
""
],
[
"von Hardenberg",
"Jost",
""
],
[
"Zappa",
"Giuseppe",
""
]
] | TITLE: RainScaleGAN: a Conditional Generative Adversarial Network for Rainfall
Downscaling
ABSTRACT: To this day, accurately simulating local-scale precipitation and reliably
reproducing its distribution remains a challenging task. The limited horizontal
resolution of Global Climate Models is among the primary factors undermining
their skill in this context. The physical mechanisms driving the onset and
development of precipitation, especially in extreme events, operate at
spatio-temporal scales smaller than those numerically resolved, thus struggling
to be captured accurately. In order to circumvent this limitation, several
downscaling approaches have been developed over the last decades to address the
discrepancy between the spatial resolution of models output and the resolution
required by local-scale applications. In this paper, we introduce RainScaleGAN,
a conditional deep convolutional Generative Adversarial Network (GAN) for
precipitation downscaling. GANs have been effectively used in image
super-resolution, an approach highly relevant for downscaling tasks.
RainScaleGAN's capabilities are tested in a perfect-model setup, where the
spatial resolution of a precipitation dataset is artificially degraded from
0.25$^{\circ}\times$0.25$^{\circ}$ to 2$^{\circ}\times$2$^\circ$, and
RainScaleGAN is used to restore it. The developed model outperforms one of the
leading precipitation downscaling method found in the literature. RainScaleGAN
not only generates a synthetic dataset featuring plausible high-resolution
spatial patterns and intensities, but also produces a precipitation
distribution with statistics closely mirroring those of the ground-truth
dataset. Given that RainScaleGAN's approach is agnostic with respect to the
underlying physics, the method has the potential to be applied to other
physical variables such as surface winds or temperature.
|
2503.13329 | Tom Burnley | Beatriz Costa-Gomes, Joel Greer, Nikolai Juraschko, James Parkhurst,
Jola Mirecka, Marjan Famili, Camila Rangel-Smith, Oliver Strickson, Alan
Lowe, Mark Basham and Tom Burnley | PERC: a suite of software tools for the curation of cryoEM data with
application to simulation, modelling and machine learning | 22 pages, 4 figures | null | null | null | cs.LG cs.CE q-bio.BM | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Ease of access to data, tools and models expedites scientific research. In
structural biology there are now numerous open repositories of experimental and
simulated datasets. Being able to easily access and utilise these is crucial
for allowing researchers to make optimal use of their research effort. The
tools presented here are useful for collating existing public cryoEM datasets
and/or creating new synthetic cryoEM datasets to aid the development of novel
data processing and interpretation algorithms. In recent years, structural
biology has seen the development of a multitude of machine-learning based
algorithms for aiding numerous steps in the processing and reconstruction of
experimental datasets and the use of these approaches has become widespread.
Developing such techniques in structural biology requires access to large
datasets which can be cumbersome to curate and unwieldy to make use of. In this
paper we present a suite of Python software packages which we collectively
refer to as PERC (profet, EMPIARreader and CAKED). These are designed to reduce
the burden which data curation places upon structural biology research. The
protein structure fetcher (profet) package allows users to conveniently
download and cleave sequences or structures from the Protein Data Bank or
Alphafold databases. EMPIARreader allows lazy loading of Electron Microscopy
Public Image Archive datasets in a machine-learning compatible structure. The
Class Aggregator for Key Electron-microscopy Data (CAKED) package is designed
to seamlessly facilitate the training of machine learning models on electron
microscopy data, including electron-cryo-microscopy-specific data augmentation
and labelling. These packages may be utilised independently or as building
blocks in workflows. All are available in open source repositories and designed
to be easily extensible to facilitate more advanced workflows if required.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 16:07:56 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Costa-Gomes",
"Beatriz",
""
],
[
"Greer",
"Joel",
""
],
[
"Juraschko",
"Nikolai",
""
],
[
"Parkhurst",
"James",
""
],
[
"Mirecka",
"Jola",
""
],
[
"Famili",
"Marjan",
""
],
[
"Rangel-Smith",
"Camila",
""
],
[
"Strickson",
"Oliver",
""
],
[
"Lowe",
"Alan",
""
],
[
"Basham",
"Mark",
""
],
[
"Burnley",
"Tom",
""
]
] | TITLE: PERC: a suite of software tools for the curation of cryoEM data with
application to simulation, modelling and machine learning
ABSTRACT: Ease of access to data, tools and models expedites scientific research. In
structural biology there are now numerous open repositories of experimental and
simulated datasets. Being able to easily access and utilise these is crucial
for allowing researchers to make optimal use of their research effort. The
tools presented here are useful for collating existing public cryoEM datasets
and/or creating new synthetic cryoEM datasets to aid the development of novel
data processing and interpretation algorithms. In recent years, structural
biology has seen the development of a multitude of machine-learning based
algorithms for aiding numerous steps in the processing and reconstruction of
experimental datasets and the use of these approaches has become widespread.
Developing such techniques in structural biology requires access to large
datasets which can be cumbersome to curate and unwieldy to make use of. In this
paper we present a suite of Python software packages which we collectively
refer to as PERC (profet, EMPIARreader and CAKED). These are designed to reduce
the burden which data curation places upon structural biology research. The
protein structure fetcher (profet) package allows users to conveniently
download and cleave sequences or structures from the Protein Data Bank or
Alphafold databases. EMPIARreader allows lazy loading of Electron Microscopy
Public Image Archive datasets in a machine-learning compatible structure. The
Class Aggregator for Key Electron-microscopy Data (CAKED) package is designed
to seamlessly facilitate the training of machine learning models on electron
microscopy data, including electron-cryo-microscopy-specific data augmentation
and labelling. These packages may be utilised independently or as building
blocks in workflows. All are available in open source repositories and designed
to be easily extensible to facilitate more advanced workflows if required.
|
2503.13330 | Ricardo Bigolin Lanfredi | Ricardo Bigolin Lanfredi, Yan Zhuang, Mark Finkelstein, Praveen
Thoppey Srinivasan Balamuralikrishna, Luke Krembs, Brandon Khoury, Arthi
Reddy, Pritam Mukherjee, Neil M. Rofsky, Ronald M. Summers | LEAVS: An LLM-based Labeler for Abdominal CT Supervision | null | null | null | null | eess.IV cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Extracting structured labels from radiology reports has been employed to
create vision models to simultaneously detect several types of abnormalities.
However, existing works focus mainly on the chest region. Few works have been
investigated on abdominal radiology reports due to more complex anatomy and a
wider range of pathologies in the abdomen. We propose LEAVS (Large language
model Extractor for Abdominal Vision Supervision). This labeler can annotate
the certainty of presence and the urgency of seven types of abnormalities for
nine abdominal organs on CT radiology reports. To ensure broad coverage, we
chose abnormalities that encompass most of the finding types from CT reports.
Our approach employs a specialized chain-of-thought prompting strategy for a
locally-run LLM using sentence extraction and multiple-choice questions in a
tree-based decision system. We demonstrate that the LLM can extract several
abnormality types across abdominal organs with an average F1 score of 0.89,
significantly outperforming competing labelers and humans. Additionally, we
show that extraction of urgency labels achieved performance comparable to human
annotations. Finally, we demonstrate that the abnormality labels contain
valuable information for training a single vision model that classifies several
organs as normal or abnormal. We release our code and structured annotations
for a public CT dataset containing over 1,000 CT volumes.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 16:09:22 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Lanfredi",
"Ricardo Bigolin",
""
],
[
"Zhuang",
"Yan",
""
],
[
"Finkelstein",
"Mark",
""
],
[
"Balamuralikrishna",
"Praveen Thoppey Srinivasan",
""
],
[
"Krembs",
"Luke",
""
],
[
"Khoury",
"Brandon",
""
],
[
"Reddy",
"Arthi",
""
],
[
"Mukherjee",
"Pritam",
""
],
[
"Rofsky",
"Neil M.",
""
],
[
"Summers",
"Ronald M.",
""
]
] | TITLE: LEAVS: An LLM-based Labeler for Abdominal CT Supervision
ABSTRACT: Extracting structured labels from radiology reports has been employed to
create vision models to simultaneously detect several types of abnormalities.
However, existing works focus mainly on the chest region. Few works have been
investigated on abdominal radiology reports due to more complex anatomy and a
wider range of pathologies in the abdomen. We propose LEAVS (Large language
model Extractor for Abdominal Vision Supervision). This labeler can annotate
the certainty of presence and the urgency of seven types of abnormalities for
nine abdominal organs on CT radiology reports. To ensure broad coverage, we
chose abnormalities that encompass most of the finding types from CT reports.
Our approach employs a specialized chain-of-thought prompting strategy for a
locally-run LLM using sentence extraction and multiple-choice questions in a
tree-based decision system. We demonstrate that the LLM can extract several
abnormality types across abdominal organs with an average F1 score of 0.89,
significantly outperforming competing labelers and humans. Additionally, we
show that extraction of urgency labels achieved performance comparable to human
annotations. Finally, we demonstrate that the abnormality labels contain
valuable information for training a single vision model that classifies several
organs as normal or abnormal. We release our code and structured annotations
for a public CT dataset containing over 1,000 CT volumes.
|
2503.13358 | Daniil Selikhanovych | Daniil Selikhanovych, David Li, Aleksei Leonov, Nikita Gushchin,
Sergei Kushneriuk, Alexander Filippov, Evgeny Burnaev, Iaroslav Koshelev,
Alexander Korotin | One-Step Residual Shifting Diffusion for Image Super-Resolution via
Distillation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion models for super-resolution (SR) produce high-quality visual
results but require expensive computational costs. Despite the development of
several methods to accelerate diffusion-based SR models, some (e.g., SinSR)
fail to produce realistic perceptual details, while others (e.g., OSEDiff) may
hallucinate non-existent structures. To overcome these issues, we present RSD,
a new distillation method for ResShift, one of the top diffusion-based SR
models. Our method is based on training the student network to produce such
images that a new fake ResShift model trained on them will coincide with the
teacher model. RSD achieves single-step restoration and outperforms the teacher
by a large margin. We show that our distillation method can surpass the other
distillation-based method for ResShift - SinSR - making it on par with
state-of-the-art diffusion-based SR distillation methods. Compared to SR
methods based on pre-trained text-to-image models, RSD produces competitive
perceptual quality, provides images with better alignment to degraded input
images, and requires fewer parameters and GPU memory. We provide experimental
results on various real-world and synthetic datasets, including RealSR,
RealSet65, DRealSR, ImageNet, and DIV2K.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 16:44:08 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Selikhanovych",
"Daniil",
""
],
[
"Li",
"David",
""
],
[
"Leonov",
"Aleksei",
""
],
[
"Gushchin",
"Nikita",
""
],
[
"Kushneriuk",
"Sergei",
""
],
[
"Filippov",
"Alexander",
""
],
[
"Burnaev",
"Evgeny",
""
],
[
"Koshelev",
"Iaroslav",
""
],
[
"Korotin",
"Alexander",
""
]
] | TITLE: One-Step Residual Shifting Diffusion for Image Super-Resolution via
Distillation
ABSTRACT: Diffusion models for super-resolution (SR) produce high-quality visual
results but require expensive computational costs. Despite the development of
several methods to accelerate diffusion-based SR models, some (e.g., SinSR)
fail to produce realistic perceptual details, while others (e.g., OSEDiff) may
hallucinate non-existent structures. To overcome these issues, we present RSD,
a new distillation method for ResShift, one of the top diffusion-based SR
models. Our method is based on training the student network to produce such
images that a new fake ResShift model trained on them will coincide with the
teacher model. RSD achieves single-step restoration and outperforms the teacher
by a large margin. We show that our distillation method can surpass the other
distillation-based method for ResShift - SinSR - making it on par with
state-of-the-art diffusion-based SR distillation methods. Compared to SR
methods based on pre-trained text-to-image models, RSD produces competitive
perceptual quality, provides images with better alignment to degraded input
images, and requires fewer parameters and GPU memory. We provide experimental
results on various real-world and synthetic datasets, including RealSR,
RealSet65, DRealSR, ImageNet, and DIV2K.
|
2503.13369 | Wan Ju Kang | Wan Ju Kang, Eunki Kim, Na Min An, Sangryul Kim, Haemin Choi, Ki Hoon
Kwak, and James Thorne | Sightation Counts: Leveraging Sighted User Feedback in Building a
BLV-aligned Dataset of Diagram Descriptions | 37 pages, 10 figures, 21 tables | null | null | null | cs.AI cs.CV cs.HC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Often, the needs and visual abilities differ between the annotator group and
the end user group. Generating detailed diagram descriptions for blind and
low-vision (BLV) users is one such challenging domain. Sighted annotators could
describe visuals with ease, but existing studies have shown that direct
generations by them are costly, bias-prone, and somewhat lacking by BLV
standards. In this study, we ask sighted individuals to assess -- rather than
produce -- diagram descriptions generated by vision-language models (VLM) that
have been guided with latent supervision via a multi-pass inference. The
sighted assessments prove effective and useful to professional educators who
are themselves BLV and teach visually impaired learners. We release Sightation,
a collection of diagram description datasets spanning 5k diagrams and 137k
samples for completion, preference, retrieval, question answering, and
reasoning training purposes and demonstrate their fine-tuning potential in
various downstream tasks.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 16:52:46 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Kang",
"Wan Ju",
""
],
[
"Kim",
"Eunki",
""
],
[
"An",
"Na Min",
""
],
[
"Kim",
"Sangryul",
""
],
[
"Choi",
"Haemin",
""
],
[
"Kwak",
"Ki Hoon",
""
],
[
"Thorne",
"James",
""
]
] | TITLE: Sightation Counts: Leveraging Sighted User Feedback in Building a
BLV-aligned Dataset of Diagram Descriptions
ABSTRACT: Often, the needs and visual abilities differ between the annotator group and
the end user group. Generating detailed diagram descriptions for blind and
low-vision (BLV) users is one such challenging domain. Sighted annotators could
describe visuals with ease, but existing studies have shown that direct
generations by them are costly, bias-prone, and somewhat lacking by BLV
standards. In this study, we ask sighted individuals to assess -- rather than
produce -- diagram descriptions generated by vision-language models (VLM) that
have been guided with latent supervision via a multi-pass inference. The
sighted assessments prove effective and useful to professional educators who
are themselves BLV and teach visually impaired learners. We release Sightation,
a collection of diagram description datasets spanning 5k diagrams and 137k
samples for completion, preference, retrieval, question answering, and
reasoning training purposes and demonstrate their fine-tuning potential in
various downstream tasks.
|
2503.13371 | Xulin Fan | Xulin Fan, Heting Gao, Ziyi Chen, Peng Chang, Mei Han, Mark
Hasegawa-Johnson | SyncDiff: Diffusion-based Talking Head Synthesis with Bottlenecked
Temporal Visual Prior for Improved Synchronization | Accepted to WACV 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Talking head synthesis, also known as speech-to-lip synthesis, reconstructs
the facial motions that align with the given audio tracks. The synthesized
videos are evaluated on mainly two aspects, lip-speech synchronization and
image fidelity. Recent studies demonstrate that GAN-based and diffusion-based
models achieve state-of-the-art (SOTA) performance on this task, with
diffusion-based models achieving superior image fidelity but experiencing lower
synchronization compared to their GAN-based counterparts. To this end, we
propose SyncDiff, a simple yet effective approach to improve diffusion-based
models using a temporal pose frame with information bottleneck and
facial-informative audio features extracted from AVHuBERT, as conditioning
input into the diffusion process. We evaluate SyncDiff on two canonical talking
head datasets, LRS2 and LRS3 for direct comparison with other SOTA models.
Experiments on LRS2/LRS3 datasets show that SyncDiff achieves a synchronization
score 27.7%/62.3% relatively higher than previous diffusion-based methods,
while preserving their high-fidelity characteristics.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 16:58:53 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Fan",
"Xulin",
""
],
[
"Gao",
"Heting",
""
],
[
"Chen",
"Ziyi",
""
],
[
"Chang",
"Peng",
""
],
[
"Han",
"Mei",
""
],
[
"Hasegawa-Johnson",
"Mark",
""
]
] | TITLE: SyncDiff: Diffusion-based Talking Head Synthesis with Bottlenecked
Temporal Visual Prior for Improved Synchronization
ABSTRACT: Talking head synthesis, also known as speech-to-lip synthesis, reconstructs
the facial motions that align with the given audio tracks. The synthesized
videos are evaluated on mainly two aspects, lip-speech synchronization and
image fidelity. Recent studies demonstrate that GAN-based and diffusion-based
models achieve state-of-the-art (SOTA) performance on this task, with
diffusion-based models achieving superior image fidelity but experiencing lower
synchronization compared to their GAN-based counterparts. To this end, we
propose SyncDiff, a simple yet effective approach to improve diffusion-based
models using a temporal pose frame with information bottleneck and
facial-informative audio features extracted from AVHuBERT, as conditioning
input into the diffusion process. We evaluate SyncDiff on two canonical talking
head datasets, LRS2 and LRS3 for direct comparison with other SOTA models.
Experiments on LRS2/LRS3 datasets show that SyncDiff achieves a synchronization
score 27.7%/62.3% relatively higher than previous diffusion-based methods,
while preserving their high-fidelity characteristics.
|
2503.13385 | Qing Zhou | Qing Zhou, Junyu Gao and Qi Wang | Scale Efficient Training for Large Datasets | Accepted by CVPR2025 | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The rapid growth of dataset scales has been a key driver in advancing deep
learning research. However, as dataset scale increases, the training process
becomes increasingly inefficient due to the presence of low-value samples,
including excessive redundant samples, overly challenging samples, and
inefficient easy samples that contribute little to model improvement.To address
this challenge, we propose Scale Efficient Training (SeTa) for large datasets,
a dynamic sample pruning approach that losslessly reduces training time. To
remove low-value samples, SeTa first performs random pruning to eliminate
redundant samples, then clusters the remaining samples according to their
learning difficulty measured by loss. Building upon this clustering, a sliding
window strategy is employed to progressively remove both overly challenging and
inefficient easy clusters following an easy-to-hard curriculum.We conduct
extensive experiments on large-scale synthetic datasets, including ToCa, SS1M,
and ST+MJ, each containing over 3 million samples.SeTa reduces training costs
by up to 50\% while maintaining or improving performance, with minimal
degradation even at 70\% cost reduction. Furthermore, experiments on various
scale real datasets across various backbones (CNNs, Transformers, and Mambas)
and diverse tasks (instruction tuning, multi-view stereo, geo-localization,
composed image retrieval, referring image segmentation) demonstrate the
powerful effectiveness and universality of our approach. Code is available at
https://github.com/mrazhou/SeTa.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:13:43 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zhou",
"Qing",
""
],
[
"Gao",
"Junyu",
""
],
[
"Wang",
"Qi",
""
]
] | TITLE: Scale Efficient Training for Large Datasets
ABSTRACT: The rapid growth of dataset scales has been a key driver in advancing deep
learning research. However, as dataset scale increases, the training process
becomes increasingly inefficient due to the presence of low-value samples,
including excessive redundant samples, overly challenging samples, and
inefficient easy samples that contribute little to model improvement.To address
this challenge, we propose Scale Efficient Training (SeTa) for large datasets,
a dynamic sample pruning approach that losslessly reduces training time. To
remove low-value samples, SeTa first performs random pruning to eliminate
redundant samples, then clusters the remaining samples according to their
learning difficulty measured by loss. Building upon this clustering, a sliding
window strategy is employed to progressively remove both overly challenging and
inefficient easy clusters following an easy-to-hard curriculum.We conduct
extensive experiments on large-scale synthetic datasets, including ToCa, SS1M,
and ST+MJ, each containing over 3 million samples.SeTa reduces training costs
by up to 50\% while maintaining or improving performance, with minimal
degradation even at 70\% cost reduction. Furthermore, experiments on various
scale real datasets across various backbones (CNNs, Transformers, and Mambas)
and diverse tasks (instruction tuning, multi-view stereo, geo-localization,
composed image retrieval, referring image segmentation) demonstrate the
powerful effectiveness and universality of our approach. Code is available at
https://github.com/mrazhou/SeTa.
|
2503.13399 | James Burgess | James Burgess, Jeffrey J Nirschl, Laura Bravo-S\'anchez, Alejandro
Lozano, Sanket Rajan Gupte, Jesus G. Galaz-Montoya, Yuhui Zhang, Yuchang Su,
Disha Bhowmik, Zachary Coman, Sarina M. Hasan, Alexandra Johannesson, William
D. Leineweber, Malvika G Nair, Ridhi Yarlagadda, Connor Zuraski, Wah Chiu,
Sarah Cohen, Jan N. Hansen, Manuel D Leonetti, Chad Liu, Emma Lundberg,
Serena Yeung-Levy | MicroVQA: A Multimodal Reasoning Benchmark for Microscopy-Based
Scientific Research | CVPR 2025 (Conference on Computer Vision and Pattern Recognition)
Project page at https://jmhb0.github.io/microvqa Benchmark at
https://huggingface.co/datasets/jmhb/microvqa | null | null | null | cs.CV cs.AI cs.CL cs.LG q-bio.CB | http://creativecommons.org/licenses/by/4.0/ | Scientific research demands sophisticated reasoning over multimodal data, a
challenge especially prevalent in biology. Despite recent advances in
multimodal large language models (MLLMs) for AI-assisted research, existing
multimodal reasoning benchmarks only target up to college-level difficulty,
while research-level benchmarks emphasize lower-level perception, falling short
of the complex multimodal reasoning needed for scientific discovery. To bridge
this gap, we introduce MicroVQA, a visual-question answering (VQA) benchmark
designed to assess three reasoning capabilities vital in research workflows:
expert image understanding, hypothesis generation, and experiment proposal.
MicroVQA consists of 1,042 multiple-choice questions (MCQs) curated by biology
experts across diverse microscopy modalities, ensuring VQA samples represent
real scientific practice. In constructing the benchmark, we find that standard
MCQ generation methods induce language shortcuts, motivating a new two-stage
pipeline: an optimized LLM prompt structures question-answer pairs into MCQs;
then, an agent-based `RefineBot' updates them to remove shortcuts. Benchmarking
on state-of-the-art MLLMs reveal a peak performance of 53\%; models with
smaller LLMs only slightly underperform top models, suggesting that
language-based reasoning is less challenging than multimodal reasoning; and
tuning with scientific articles enhances performance. Expert analysis of
chain-of-thought responses shows that perception errors are the most frequent,
followed by knowledge errors and then overgeneralization errors. These insights
highlight the challenges in multimodal scientific reasoning, showing MicroVQA
is a valuable resource advancing AI-driven biomedical research. MicroVQA is
available at https://huggingface.co/datasets/jmhb/microvqa, and project page at
https://jmhb0.github.io/microvqa.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:33:10 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Burgess",
"James",
""
],
[
"Nirschl",
"Jeffrey J",
""
],
[
"Bravo-Sánchez",
"Laura",
""
],
[
"Lozano",
"Alejandro",
""
],
[
"Gupte",
"Sanket Rajan",
""
],
[
"Galaz-Montoya",
"Jesus G.",
""
],
[
"Zhang",
"Yuhui",
""
],
[
"Su",
"Yuchang",
""
],
[
"Bhowmik",
"Disha",
""
],
[
"Coman",
"Zachary",
""
],
[
"Hasan",
"Sarina M.",
""
],
[
"Johannesson",
"Alexandra",
""
],
[
"Leineweber",
"William D.",
""
],
[
"Nair",
"Malvika G",
""
],
[
"Yarlagadda",
"Ridhi",
""
],
[
"Zuraski",
"Connor",
""
],
[
"Chiu",
"Wah",
""
],
[
"Cohen",
"Sarah",
""
],
[
"Hansen",
"Jan N.",
""
],
[
"Leonetti",
"Manuel D",
""
],
[
"Liu",
"Chad",
""
],
[
"Lundberg",
"Emma",
""
],
[
"Yeung-Levy",
"Serena",
""
]
] | TITLE: MicroVQA: A Multimodal Reasoning Benchmark for Microscopy-Based
Scientific Research
ABSTRACT: Scientific research demands sophisticated reasoning over multimodal data, a
challenge especially prevalent in biology. Despite recent advances in
multimodal large language models (MLLMs) for AI-assisted research, existing
multimodal reasoning benchmarks only target up to college-level difficulty,
while research-level benchmarks emphasize lower-level perception, falling short
of the complex multimodal reasoning needed for scientific discovery. To bridge
this gap, we introduce MicroVQA, a visual-question answering (VQA) benchmark
designed to assess three reasoning capabilities vital in research workflows:
expert image understanding, hypothesis generation, and experiment proposal.
MicroVQA consists of 1,042 multiple-choice questions (MCQs) curated by biology
experts across diverse microscopy modalities, ensuring VQA samples represent
real scientific practice. In constructing the benchmark, we find that standard
MCQ generation methods induce language shortcuts, motivating a new two-stage
pipeline: an optimized LLM prompt structures question-answer pairs into MCQs;
then, an agent-based `RefineBot' updates them to remove shortcuts. Benchmarking
on state-of-the-art MLLMs reveal a peak performance of 53\%; models with
smaller LLMs only slightly underperform top models, suggesting that
language-based reasoning is less challenging than multimodal reasoning; and
tuning with scientific articles enhances performance. Expert analysis of
chain-of-thought responses shows that perception errors are the most frequent,
followed by knowledge errors and then overgeneralization errors. These insights
highlight the challenges in multimodal scientific reasoning, showing MicroVQA
is a valuable resource advancing AI-driven biomedical research. MicroVQA is
available at https://huggingface.co/datasets/jmhb/microvqa, and project page at
https://jmhb0.github.io/microvqa.
|
2503.13400 | Qi Zhang | Qi Zhang, Xiuyuan Chen, Ziyi He, Kun Wang, Lianming Wu, Hongxing Shen,
and Jianqi Sun | U2AD: Uncertainty-based Unsupervised Anomaly Detection Framework for
Detecting T2 Hyperintensity in MRI Spinal Cord | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | T2 hyperintensities in spinal cord MR images are crucial biomarkers for
conditions such as degenerative cervical myelopathy. However, current clinical
diagnoses primarily rely on manual evaluation. Deep learning methods have shown
promise in lesion detection, but most supervised approaches are heavily
dependent on large, annotated datasets. Unsupervised anomaly detection (UAD)
offers a compelling alternative by eliminating the need for abnormal data
annotations. However, existing UAD methods rely on curated normal datasets and
their performance frequently deteriorates when applied to clinical datasets due
to domain shifts. We propose an Uncertainty-based Unsupervised Anomaly
Detection framework, termed U2AD, to address these limitations. Unlike
traditional methods, U2AD is designed to be trained and tested within the same
clinical dataset, following a "mask-and-reconstruction" paradigm built on a
Vision Transformer-based architecture. We introduce an uncertainty-guided
masking strategy to resolve task conflicts between normal reconstruction and
anomaly detection to achieve an optimal balance. Specifically, we employ a
Monte-Carlo sampling technique to estimate reconstruction uncertainty mappings
during training. By iteratively optimizing reconstruction training under the
guidance of both epistemic and aleatoric uncertainty, U2AD reduces overall
reconstruction variance while emphasizing regions. Experimental results
demonstrate that U2AD outperforms existing supervised and unsupervised methods
in patient-level identification and segment-level localization tasks. This
framework establishes a new benchmark for incorporating uncertainty guidance
into UAD, highlighting its clinical utility in addressing domain shifts and
task conflicts in medical image anomaly detection. Our code is available:
https://github.com/zhibaishouheilab/U2AD
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:33:32 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Zhang",
"Qi",
""
],
[
"Chen",
"Xiuyuan",
""
],
[
"He",
"Ziyi",
""
],
[
"Wang",
"Kun",
""
],
[
"Wu",
"Lianming",
""
],
[
"Shen",
"Hongxing",
""
],
[
"Sun",
"Jianqi",
""
]
] | TITLE: U2AD: Uncertainty-based Unsupervised Anomaly Detection Framework for
Detecting T2 Hyperintensity in MRI Spinal Cord
ABSTRACT: T2 hyperintensities in spinal cord MR images are crucial biomarkers for
conditions such as degenerative cervical myelopathy. However, current clinical
diagnoses primarily rely on manual evaluation. Deep learning methods have shown
promise in lesion detection, but most supervised approaches are heavily
dependent on large, annotated datasets. Unsupervised anomaly detection (UAD)
offers a compelling alternative by eliminating the need for abnormal data
annotations. However, existing UAD methods rely on curated normal datasets and
their performance frequently deteriorates when applied to clinical datasets due
to domain shifts. We propose an Uncertainty-based Unsupervised Anomaly
Detection framework, termed U2AD, to address these limitations. Unlike
traditional methods, U2AD is designed to be trained and tested within the same
clinical dataset, following a "mask-and-reconstruction" paradigm built on a
Vision Transformer-based architecture. We introduce an uncertainty-guided
masking strategy to resolve task conflicts between normal reconstruction and
anomaly detection to achieve an optimal balance. Specifically, we employ a
Monte-Carlo sampling technique to estimate reconstruction uncertainty mappings
during training. By iteratively optimizing reconstruction training under the
guidance of both epistemic and aleatoric uncertainty, U2AD reduces overall
reconstruction variance while emphasizing regions. Experimental results
demonstrate that U2AD outperforms existing supervised and unsupervised methods
in patient-level identification and segment-level localization tasks. This
framework establishes a new benchmark for incorporating uncertainty guidance
into UAD, highlighting its clinical utility in addressing domain shifts and
task conflicts in medical image anomaly detection. Our code is available:
https://github.com/zhibaishouheilab/U2AD
|
2503.13404 | Xubo Yue | Cheoljoon Jeong, Xubo Yue, Seokhyun Chung | Fed-Joint: Joint Modeling of Nonlinear Degradation Signals and Failure
Events for Remaining Useful Life Prediction using Federated Learning | null | null | null | null | cs.AI cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many failure mechanisms of machinery are closely related to the behavior of
condition monitoring (CM) signals. To achieve a cost-effective preventive
maintenance strategy, accurate remaining useful life (RUL) prediction based on
the signals is of paramount importance. However, the CM signals are often
recorded at different factories and production lines, with limited amounts of
data. Unfortunately, these datasets have rarely been shared between the sites
due to data confidentiality and ownership issues, a lack of computing and
storage power, and high communication costs associated with data transfer
between sites and a data center. Another challenge in real applications is that
the CM signals are often not explicitly specified \textit{a priori}, meaning
that existing methods, which often usually a parametric form, may not be
applicable. To address these challenges, we propose a new prognostic framework
for RUL prediction using the joint modeling of nonlinear degradation signals
and time-to-failure data within a federated learning scheme. The proposed
method constructs a nonparametric degradation model using a federated
multi-output Gaussian process and then employs a federated survival model to
predict failure times and probabilities for in-service machinery. The
superiority of the proposed method over other alternatives is demonstrated
through comprehensive simulation studies and a case study using turbofan engine
degradation signal data that include run-to-failure events.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:34:34 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Jeong",
"Cheoljoon",
""
],
[
"Yue",
"Xubo",
""
],
[
"Chung",
"Seokhyun",
""
]
] | TITLE: Fed-Joint: Joint Modeling of Nonlinear Degradation Signals and Failure
Events for Remaining Useful Life Prediction using Federated Learning
ABSTRACT: Many failure mechanisms of machinery are closely related to the behavior of
condition monitoring (CM) signals. To achieve a cost-effective preventive
maintenance strategy, accurate remaining useful life (RUL) prediction based on
the signals is of paramount importance. However, the CM signals are often
recorded at different factories and production lines, with limited amounts of
data. Unfortunately, these datasets have rarely been shared between the sites
due to data confidentiality and ownership issues, a lack of computing and
storage power, and high communication costs associated with data transfer
between sites and a data center. Another challenge in real applications is that
the CM signals are often not explicitly specified \textit{a priori}, meaning
that existing methods, which often usually a parametric form, may not be
applicable. To address these challenges, we propose a new prognostic framework
for RUL prediction using the joint modeling of nonlinear degradation signals
and time-to-failure data within a federated learning scheme. The proposed
method constructs a nonparametric degradation model using a federated
multi-output Gaussian process and then employs a federated survival model to
predict failure times and probabilities for in-service machinery. The
superiority of the proposed method over other alternatives is demonstrated
through comprehensive simulation studies and a case study using turbofan engine
degradation signal data that include run-to-failure events.
|
2503.13409 | Guillaume Lagarde | Gabriel Bathie and Guillaume Lagarde | A $(1+\epsilon)$-Approximation for Ultrametric Embedding in Subquadratic
Time | Extended version of AAAI 2025 | null | null | null | cs.DS | http://creativecommons.org/licenses/by/4.0/ | Efficiently computing accurate representations of high-dimensional data is
essential for data analysis and unsupervised learning. Dendrograms, also known
as ultrametrics, are widely used representations that preserve hierarchical
relationships within the data. However, popular methods for computing them,
such as linkage algorithms, suffer from quadratic time and space complexity,
making them impractical for large datasets.
The "best ultrametric embedding" (a.k.a. "best ultrametric fit") problem,
which aims to find the ultrametric that best preserves the distances between
points in the original data, is known to require at least quadratic time for an
exact solution.
Recent work has focused on improving scalability by approximating optimal
solutions in subquadratic time, resulting in a $(\sqrt{2} +
\epsilon)$-approximation (Cohen-Addad, de Joannis de Verclos and Lagarde,
2021).
In this paper, we present the first subquadratic algorithm that achieves
arbitrarily precise approximations of the optimal ultrametric embedding.
Specifically, we provide an algorithm that, for any $c \geq 1$, outputs a
$c$-approximation of the best ultrametric in time $\tilde{O}(n^{1 + 1/c})$. In
particular, for any fixed $\epsilon > 0$, the algorithm computes a
$(1+\epsilon)$-approximation in time $\tilde{O}(n^{2 - \epsilon +
o(\epsilon^2)})$.
Experimental results show that our algorithm improves upon previous methods
in terms of approximation quality while maintaining comparable running times.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:38:37 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Bathie",
"Gabriel",
""
],
[
"Lagarde",
"Guillaume",
""
]
] | TITLE: A $(1+\epsilon)$-Approximation for Ultrametric Embedding in Subquadratic
Time
ABSTRACT: Efficiently computing accurate representations of high-dimensional data is
essential for data analysis and unsupervised learning. Dendrograms, also known
as ultrametrics, are widely used representations that preserve hierarchical
relationships within the data. However, popular methods for computing them,
such as linkage algorithms, suffer from quadratic time and space complexity,
making them impractical for large datasets.
The "best ultrametric embedding" (a.k.a. "best ultrametric fit") problem,
which aims to find the ultrametric that best preserves the distances between
points in the original data, is known to require at least quadratic time for an
exact solution.
Recent work has focused on improving scalability by approximating optimal
solutions in subquadratic time, resulting in a $(\sqrt{2} +
\epsilon)$-approximation (Cohen-Addad, de Joannis de Verclos and Lagarde,
2021).
In this paper, we present the first subquadratic algorithm that achieves
arbitrarily precise approximations of the optimal ultrametric embedding.
Specifically, we provide an algorithm that, for any $c \geq 1$, outputs a
$c$-approximation of the best ultrametric in time $\tilde{O}(n^{1 + 1/c})$. In
particular, for any fixed $\epsilon > 0$, the algorithm computes a
$(1+\epsilon)$-approximation in time $\tilde{O}(n^{2 - \epsilon +
o(\epsilon^2)})$.
Experimental results show that our algorithm improves upon previous methods
in terms of approximation quality while maintaining comparable running times.
|
2503.13419 | Khaza Anuarul Hoque | Ripan Kumar Kundu, Matthew Denton, Genova Mongalo, Prasad Calyam,
Khaza Anuarul Hoque | Securing Virtual Reality Experiences: Unveiling and Tackling
Cybersickness Attacks with Explainable AI | This work has been submitted to the IEEE for possible publication | null | null | null | cs.CR cs.AI cs.ET cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The synergy between virtual reality (VR) and artificial intelligence (AI),
specifically deep learning (DL)-based cybersickness detection models, has
ushered in unprecedented advancements in immersive experiences by automatically
detecting cybersickness severity and adaptively various mitigation techniques,
offering a smooth and comfortable VR experience. While this DL-enabled
cybersickness detection method provides promising solutions for enhancing user
experiences, it also introduces new risks since these models are vulnerable to
adversarial attacks; a small perturbation of the input data that is visually
undetectable to human observers can fool the cybersickness detection model and
trigger unexpected mitigation, thus disrupting user immersive experiences (UIX)
and even posing safety risks. In this paper, we present a new type of VR
attack, i.e., a cybersickness attack, which successfully stops the triggering
of cybersickness mitigation by fooling DL-based cybersickness detection models
and dramatically hinders the UIX. Next, we propose a novel explainable
artificial intelligence (XAI)-guided cybersickness attack detection framework
to detect such attacks in VR to ensure UIX and a comfortable VR experience. We
evaluate the proposed attack and the detection framework using two
state-of-the-art open-source VR cybersickness datasets: Simulation 2021 and
Gameplay dataset. Finally, to verify the effectiveness of our proposed method,
we implement the attack and the XAI-based detection using a testbed with a
custom-built VR roller coaster simulation with an HTC Vive Pro Eye headset and
perform a user study. Our study shows that such an attack can dramatically
hinder the UIX. However, our proposed XAI-guided cybersickness attack detection
can successfully detect cybersickness attacks and trigger the proper
mitigation, effectively reducing VR cybersickness.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:49:51 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Kundu",
"Ripan Kumar",
""
],
[
"Denton",
"Matthew",
""
],
[
"Mongalo",
"Genova",
""
],
[
"Calyam",
"Prasad",
""
],
[
"Hoque",
"Khaza Anuarul",
""
]
] | TITLE: Securing Virtual Reality Experiences: Unveiling and Tackling
Cybersickness Attacks with Explainable AI
ABSTRACT: The synergy between virtual reality (VR) and artificial intelligence (AI),
specifically deep learning (DL)-based cybersickness detection models, has
ushered in unprecedented advancements in immersive experiences by automatically
detecting cybersickness severity and adaptively various mitigation techniques,
offering a smooth and comfortable VR experience. While this DL-enabled
cybersickness detection method provides promising solutions for enhancing user
experiences, it also introduces new risks since these models are vulnerable to
adversarial attacks; a small perturbation of the input data that is visually
undetectable to human observers can fool the cybersickness detection model and
trigger unexpected mitigation, thus disrupting user immersive experiences (UIX)
and even posing safety risks. In this paper, we present a new type of VR
attack, i.e., a cybersickness attack, which successfully stops the triggering
of cybersickness mitigation by fooling DL-based cybersickness detection models
and dramatically hinders the UIX. Next, we propose a novel explainable
artificial intelligence (XAI)-guided cybersickness attack detection framework
to detect such attacks in VR to ensure UIX and a comfortable VR experience. We
evaluate the proposed attack and the detection framework using two
state-of-the-art open-source VR cybersickness datasets: Simulation 2021 and
Gameplay dataset. Finally, to verify the effectiveness of our proposed method,
we implement the attack and the XAI-based detection using a testbed with a
custom-built VR roller coaster simulation with an HTC Vive Pro Eye headset and
perform a user study. Our study shows that such an attack can dramatically
hinder the UIX. However, our proposed XAI-guided cybersickness attack detection
can successfully detect cybersickness attacks and trigger the proper
mitigation, effectively reducing VR cybersickness.
|
2503.13424 | Xinyu Lian | Xinyu Lian, Zichao Yu, Ruiming Liang, Yitong Wang, Li Ray Luo, Kaixu
Chen, Yuanzhen Zhou, Qihong Tang, Xudong Xu, Zhaoyang Lyu, Bo Dai, Jiangmiao
Pang | Infinite Mobility: Scalable High-Fidelity Synthesis of Articulated
Objects via Procedural Generation | Project page: https://infinite-mobility.github.io 10 pages,12 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Large-scale articulated objects with high quality are desperately needed for
multiple tasks related to embodied AI. Most existing methods for creating
articulated objects are either data-driven or simulation based, which are
limited by the scale and quality of the training data or the fidelity and heavy
labour of the simulation. In this paper, we propose Infinite Mobility, a novel
method for synthesizing high-fidelity articulated objects through procedural
generation. User study and quantitative evaluation demonstrate that our method
can produce results that excel current state-of-the-art methods and are
comparable to human-annotated datasets in both physics property and mesh
quality. Furthermore, we show that our synthetic data can be used as training
data for generative models, enabling next-step scaling up. Code is available at
https://github.com/Intern-Nexus/Infinite-Mobility
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:53:56 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Lian",
"Xinyu",
""
],
[
"Yu",
"Zichao",
""
],
[
"Liang",
"Ruiming",
""
],
[
"Wang",
"Yitong",
""
],
[
"Luo",
"Li Ray",
""
],
[
"Chen",
"Kaixu",
""
],
[
"Zhou",
"Yuanzhen",
""
],
[
"Tang",
"Qihong",
""
],
[
"Xu",
"Xudong",
""
],
[
"Lyu",
"Zhaoyang",
""
],
[
"Dai",
"Bo",
""
],
[
"Pang",
"Jiangmiao",
""
]
] | TITLE: Infinite Mobility: Scalable High-Fidelity Synthesis of Articulated
Objects via Procedural Generation
ABSTRACT: Large-scale articulated objects with high quality are desperately needed for
multiple tasks related to embodied AI. Most existing methods for creating
articulated objects are either data-driven or simulation based, which are
limited by the scale and quality of the training data or the fidelity and heavy
labour of the simulation. In this paper, we propose Infinite Mobility, a novel
method for synthesizing high-fidelity articulated objects through procedural
generation. User study and quantitative evaluation demonstrate that our method
can produce results that excel current state-of-the-art methods and are
comparable to human-annotated datasets in both physics property and mesh
quality. Furthermore, we show that our synthetic data can be used as training
data for generative models, enabling next-step scaling up. Code is available at
https://github.com/Intern-Nexus/Infinite-Mobility
|
2503.13430 | Thomas Monninger | Thomas Monninger, Md Zafar Anwar, Stanislaw Antol, Steffen Staab,
Sihao Ding | AugMapNet: Improving Spatial Latent Structure via BEV Grid Augmentation
for Enhanced Vectorized Online HD Map Construction | null | null | null | null | cs.CV cs.AI cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Autonomous driving requires an understanding of the infrastructure elements,
such as lanes and crosswalks. To navigate safely, this understanding must be
derived from sensor data in real-time and needs to be represented in vectorized
form. Learned Bird's-Eye View (BEV) encoders are commonly used to combine a set
of camera images from multiple views into one joint latent BEV grid.
Traditionally, from this latent space, an intermediate raster map is predicted,
providing dense spatial supervision but requiring post-processing into the
desired vectorized form. More recent models directly derive infrastructure
elements as polylines using vectorized map decoders, providing instance-level
information. Our approach, Augmentation Map Network (AugMapNet), proposes
latent BEV grid augmentation, a novel technique that significantly enhances the
latent BEV representation. AugMapNet combines vector decoding and dense spatial
supervision more effectively than existing architectures while remaining as
straightforward to integrate and as generic as auxiliary supervision.
Experiments on nuScenes and Argoverse2 datasets demonstrate significant
improvements in vectorized map prediction performance up to 13.3% over the
StreamMapNet baseline on 60m range and greater improvements on larger ranges.
We confirm transferability by applying our method to another baseline and find
similar improvements. A detailed analysis of the latent BEV grid confirms a
more structured latent space of AugMapNet and shows the value of our novel
concept beyond pure performance improvement. The code will be released soon.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:55:32 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Monninger",
"Thomas",
""
],
[
"Anwar",
"Md Zafar",
""
],
[
"Antol",
"Stanislaw",
""
],
[
"Staab",
"Steffen",
""
],
[
"Ding",
"Sihao",
""
]
] | TITLE: AugMapNet: Improving Spatial Latent Structure via BEV Grid Augmentation
for Enhanced Vectorized Online HD Map Construction
ABSTRACT: Autonomous driving requires an understanding of the infrastructure elements,
such as lanes and crosswalks. To navigate safely, this understanding must be
derived from sensor data in real-time and needs to be represented in vectorized
form. Learned Bird's-Eye View (BEV) encoders are commonly used to combine a set
of camera images from multiple views into one joint latent BEV grid.
Traditionally, from this latent space, an intermediate raster map is predicted,
providing dense spatial supervision but requiring post-processing into the
desired vectorized form. More recent models directly derive infrastructure
elements as polylines using vectorized map decoders, providing instance-level
information. Our approach, Augmentation Map Network (AugMapNet), proposes
latent BEV grid augmentation, a novel technique that significantly enhances the
latent BEV representation. AugMapNet combines vector decoding and dense spatial
supervision more effectively than existing architectures while remaining as
straightforward to integrate and as generic as auxiliary supervision.
Experiments on nuScenes and Argoverse2 datasets demonstrate significant
improvements in vectorized map prediction performance up to 13.3% over the
StreamMapNet baseline on 60m range and greater improvements on larger ranges.
We confirm transferability by applying our method to another baseline and find
similar improvements. A detailed analysis of the latent BEV grid confirms a
more structured latent space of AugMapNet and shows the value of our novel
concept beyond pure performance improvement. The code will be released soon.
|
2503.13435 | Ling Yang | Ling Yang, Kaixin Zhu, Juanxi Tian, Bohan Zeng, Mingbao Lin, Hongjuan
Pei, Wentao Zhang, Shuicheng Yan | WideRange4D: Enabling High-Quality 4D Reconstruction with Wide-Range
Movements and Scenes | Project: https://github.com/Gen-Verse/WideRange4D | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid development of 3D reconstruction technology, research in 4D
reconstruction is also advancing, existing 4D reconstruction methods can
generate high-quality 4D scenes. However, due to the challenges in acquiring
multi-view video data, the current 4D reconstruction benchmarks mainly display
actions performed in place, such as dancing, within limited scenarios. In
practical scenarios, many scenes involve wide-range spatial movements,
highlighting the limitations of existing 4D reconstruction datasets.
Additionally, existing 4D reconstruction methods rely on deformation fields to
estimate the dynamics of 3D objects, but deformation fields struggle with
wide-range spatial movements, which limits the ability to achieve high-quality
4D scene reconstruction with wide-range spatial movements. In this paper, we
focus on 4D scene reconstruction with significant object spatial movements and
propose a novel 4D reconstruction benchmark, WideRange4D. This benchmark
includes rich 4D scene data with large spatial variations, allowing for a more
comprehensive evaluation of the generation capabilities of 4D generation
methods. Furthermore, we introduce a new 4D reconstruction method, Progress4D,
which generates stable and high-quality 4D results across various complex 4D
scene reconstruction tasks. We conduct both quantitative and qualitative
comparison experiments on WideRange4D, showing that our Progress4D outperforms
existing state-of-the-art 4D reconstruction methods. Project:
https://github.com/Gen-Verse/WideRange4D
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:58:18 GMT"
}
] | 2025-03-18T00:00:00 | [
[
"Yang",
"Ling",
""
],
[
"Zhu",
"Kaixin",
""
],
[
"Tian",
"Juanxi",
""
],
[
"Zeng",
"Bohan",
""
],
[
"Lin",
"Mingbao",
""
],
[
"Pei",
"Hongjuan",
""
],
[
"Zhang",
"Wentao",
""
],
[
"Yan",
"Shuicheng",
""
]
] | TITLE: WideRange4D: Enabling High-Quality 4D Reconstruction with Wide-Range
Movements and Scenes
ABSTRACT: With the rapid development of 3D reconstruction technology, research in 4D
reconstruction is also advancing, existing 4D reconstruction methods can
generate high-quality 4D scenes. However, due to the challenges in acquiring
multi-view video data, the current 4D reconstruction benchmarks mainly display
actions performed in place, such as dancing, within limited scenarios. In
practical scenarios, many scenes involve wide-range spatial movements,
highlighting the limitations of existing 4D reconstruction datasets.
Additionally, existing 4D reconstruction methods rely on deformation fields to
estimate the dynamics of 3D objects, but deformation fields struggle with
wide-range spatial movements, which limits the ability to achieve high-quality
4D scene reconstruction with wide-range spatial movements. In this paper, we
focus on 4D scene reconstruction with significant object spatial movements and
propose a novel 4D reconstruction benchmark, WideRange4D. This benchmark
includes rich 4D scene data with large spatial variations, allowing for a more
comprehensive evaluation of the generation capabilities of 4D generation
methods. Furthermore, we introduce a new 4D reconstruction method, Progress4D,
which generates stable and high-quality 4D results across various complex 4D
scene reconstruction tasks. We conduct both quantitative and qualitative
comparison experiments on WideRange4D, showing that our Progress4D outperforms
existing state-of-the-art 4D reconstruction methods. Project:
https://github.com/Gen-Verse/WideRange4D
|
2209.04517 | Jola Mirecka | Marjan Famili, Jola Mirecka, Camila Rangel Smith, Anna Kota\'nska,
Nikolai Juraschko, Beatriz Costa-Gomes, Colin M. Palmer, Jeyan Thiyagalingam,
Tom Burnley, Mark Basham, Alan R. Lowe | Affinity-VAE: incorporating prior knowledge in representation learning
from scientific images | null | null | null | null | cs.CV cs.LG q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning compact and interpretable representations of data is a critical
challenge in scientific image analysis. Here, we introduce Affinity-VAE, a
generative model that enables us to impose our scientific intuition about the
similarity of instances in the dataset on the learned representation during
training. We demonstrate the utility of the approach in the scientific domain
of cryo-electron tomography (cryo-ET) where a significant current challenge is
to identify similar molecules within a noisy and low contrast tomographic image
volume. This task is distinct from classification in that, at inference time,
it is unknown whether an instance is part of the training set or not. We
trained affinity-VAE using prior knowledge of protein structure to inform the
latent space. Our model is able to create rotationally-invariant,
morphologically homogeneous clusters in the latent representation, with
improved cluster separation compared to other approaches. It achieves
competitive performance on protein classification with the added benefit of
disentangling object pose, structural similarity and an interpretable latent
representation. In the context of cryo-ET data, affinity-VAE captures the
orientation of identified proteins in 3D which can be used as a prior for
subsequent scientific experiments. Extracting physical principles from a
trained network is of significant importance in scientific imaging where a
ground truth training set is not always feasible.
| [
{
"version": "v1",
"created": "Fri, 9 Sep 2022 20:39:22 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 16:34:24 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Famili",
"Marjan",
""
],
[
"Mirecka",
"Jola",
""
],
[
"Smith",
"Camila Rangel",
""
],
[
"Kotańska",
"Anna",
""
],
[
"Juraschko",
"Nikolai",
""
],
[
"Costa-Gomes",
"Beatriz",
""
],
[
"Palmer",
"Colin M.",
""
],
[
"Thiyagalingam",
"Jeyan",
""
],
[
"Burnley",
"Tom",
""
],
[
"Basham",
"Mark",
""
],
[
"Lowe",
"Alan R.",
""
]
] | TITLE: Affinity-VAE: incorporating prior knowledge in representation learning
from scientific images
ABSTRACT: Learning compact and interpretable representations of data is a critical
challenge in scientific image analysis. Here, we introduce Affinity-VAE, a
generative model that enables us to impose our scientific intuition about the
similarity of instances in the dataset on the learned representation during
training. We demonstrate the utility of the approach in the scientific domain
of cryo-electron tomography (cryo-ET) where a significant current challenge is
to identify similar molecules within a noisy and low contrast tomographic image
volume. This task is distinct from classification in that, at inference time,
it is unknown whether an instance is part of the training set or not. We
trained affinity-VAE using prior knowledge of protein structure to inform the
latent space. Our model is able to create rotationally-invariant,
morphologically homogeneous clusters in the latent representation, with
improved cluster separation compared to other approaches. It achieves
competitive performance on protein classification with the added benefit of
disentangling object pose, structural similarity and an interpretable latent
representation. In the context of cryo-ET data, affinity-VAE captures the
orientation of identified proteins in 3D which can be used as a prior for
subsequent scientific experiments. Extracting physical principles from a
trained network is of significant importance in scientific imaging where a
ground truth training set is not always feasible.
|
2210.14485 | Tongkun Liu | Tongkun Liu, Bing Li, Zhuo Zhao, Xiao Du, Bingke Jiang, Leqi Geng | Reconstruction from edge image combined with color and gradient
difference for industrial surface anomaly detection | 11 pages, 8 figures | null | 10.1016/j.aei.2024.103064 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reconstruction-based methods are widely explored in industrial visual anomaly
detection. Such methods commonly require the model to well reconstruct the
normal patterns but fail in the anomalies, and thus the anomalies can be
detected by evaluating the reconstruction errors. However, in practice, it's
usually difficult to control the generalization boundary of the model. The
model with an overly strong generalization capability can even well reconstruct
the abnormal regions, making them less distinguishable, while the model with a
poor generalization capability can not reconstruct those changeable
high-frequency components in the normal regions, which ultimately leads to
false positives. To tackle the above issue, we propose a new reconstruction
network where we reconstruct the original RGB image from its gray value edges
(EdgRec). Specifically, this is achieved by an UNet-type denoising autoencoder
with skip connections. The input edge and skip connections can well preserve
the high-frequency information in the original image. Meanwhile, the proposed
restoration task can force the network to memorize the normal low-frequency and
color information. Besides, the denoising design can prevent the model from
directly copying the original high-frequent components. To evaluate the
anomalies, we further propose a new interpretable hand-crafted evaluation
function that considers both the color and gradient differences. Our method
achieves competitive results on the challenging benchmark MVTec AD (97.8\% for
detection and 97.7\% for localization, AUROC). In addition, we conduct
experiments on the MVTec 3D-AD dataset and show convincing results using RGB
images only. Our code will be available at
https://github.com/liutongkun/EdgRec.
| [
{
"version": "v1",
"created": "Wed, 26 Oct 2022 05:21:43 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Liu",
"Tongkun",
""
],
[
"Li",
"Bing",
""
],
[
"Zhao",
"Zhuo",
""
],
[
"Du",
"Xiao",
""
],
[
"Jiang",
"Bingke",
""
],
[
"Geng",
"Leqi",
""
]
] | TITLE: Reconstruction from edge image combined with color and gradient
difference for industrial surface anomaly detection
ABSTRACT: Reconstruction-based methods are widely explored in industrial visual anomaly
detection. Such methods commonly require the model to well reconstruct the
normal patterns but fail in the anomalies, and thus the anomalies can be
detected by evaluating the reconstruction errors. However, in practice, it's
usually difficult to control the generalization boundary of the model. The
model with an overly strong generalization capability can even well reconstruct
the abnormal regions, making them less distinguishable, while the model with a
poor generalization capability can not reconstruct those changeable
high-frequency components in the normal regions, which ultimately leads to
false positives. To tackle the above issue, we propose a new reconstruction
network where we reconstruct the original RGB image from its gray value edges
(EdgRec). Specifically, this is achieved by an UNet-type denoising autoencoder
with skip connections. The input edge and skip connections can well preserve
the high-frequency information in the original image. Meanwhile, the proposed
restoration task can force the network to memorize the normal low-frequency and
color information. Besides, the denoising design can prevent the model from
directly copying the original high-frequent components. To evaluate the
anomalies, we further propose a new interpretable hand-crafted evaluation
function that considers both the color and gradient differences. Our method
achieves competitive results on the challenging benchmark MVTec AD (97.8\% for
detection and 97.7\% for localization, AUROC). In addition, we conduct
experiments on the MVTec 3D-AD dataset and show convincing results using RGB
images only. Our code will be available at
https://github.com/liutongkun/EdgRec.
|
2303.03470 | R. Spencer Hallyburton | R. Spencer Hallyburton, Qingzhao Zhang, Z. Morley Mao, Michael Reiter,
Miroslav Pajic | What Would Trojans Do? Exploiting Partial-Information Vulnerabilities in
Autonomous Vehicle Sensing | null | null | null | null | cs.CR cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Safety-critical sensors in autonomous vehicles (AVs) form an essential part
of the vehicle's trusted computing base (TCB), yet they are highly susceptible
to attacks. Alarmingly, Tier 1 manufacturers have already exposed
vulnerabilities to attacks introducing Trojans that can stealthily alter sensor
outputs. We analyze the feasible capability and safety-critical outcomes of an
attack on sensing at a cyber level. To further address these threats, we design
realistic attacks in AV simulators and real-world datasets under two practical
constraints: attackers (1) possess only partial information and (2) are
constrained by data structures that maintain sensor integrity.Examining the
role of camera and LiDAR in multi-sensor AVs, we find that attacks targeting
only the camera have minimal safety impact due to the sensor fusion system's
strong reliance on 3D data from LiDAR. This reliance makes LiDAR-based attacks
especially detrimental to safety. To mitigate the vulnerabilities, we introduce
security-aware sensor fusion incorporating (1) a probabilistic data-asymmetry
monitor and (2) a scalable track-to-track fusion of 3D LiDAR and monocular
detections (T2T-3DLM). We demonstrate that these methods significantly diminish
attack success rate.
| [
{
"version": "v1",
"created": "Mon, 6 Mar 2023 19:52:41 GMT"
},
{
"version": "v2",
"created": "Sun, 23 Apr 2023 11:41:26 GMT"
},
{
"version": "v3",
"created": "Fri, 8 Dec 2023 10:58:13 GMT"
},
{
"version": "v4",
"created": "Thu, 13 Mar 2025 20:57:41 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Hallyburton",
"R. Spencer",
""
],
[
"Zhang",
"Qingzhao",
""
],
[
"Mao",
"Z. Morley",
""
],
[
"Reiter",
"Michael",
""
],
[
"Pajic",
"Miroslav",
""
]
] | TITLE: What Would Trojans Do? Exploiting Partial-Information Vulnerabilities in
Autonomous Vehicle Sensing
ABSTRACT: Safety-critical sensors in autonomous vehicles (AVs) form an essential part
of the vehicle's trusted computing base (TCB), yet they are highly susceptible
to attacks. Alarmingly, Tier 1 manufacturers have already exposed
vulnerabilities to attacks introducing Trojans that can stealthily alter sensor
outputs. We analyze the feasible capability and safety-critical outcomes of an
attack on sensing at a cyber level. To further address these threats, we design
realistic attacks in AV simulators and real-world datasets under two practical
constraints: attackers (1) possess only partial information and (2) are
constrained by data structures that maintain sensor integrity.Examining the
role of camera and LiDAR in multi-sensor AVs, we find that attacks targeting
only the camera have minimal safety impact due to the sensor fusion system's
strong reliance on 3D data from LiDAR. This reliance makes LiDAR-based attacks
especially detrimental to safety. To mitigate the vulnerabilities, we introduce
security-aware sensor fusion incorporating (1) a probabilistic data-asymmetry
monitor and (2) a scalable track-to-track fusion of 3D LiDAR and monocular
detections (T2T-3DLM). We demonstrate that these methods significantly diminish
attack success rate.
|
2303.17448 | Weiming Li | Weiming Li, Xueqian Wang, Gang Li, Baocheng Geng, Pramod K. Varshney | NN-Copula-CD: A Copula-Guided Interpretable Neural Network for Change
Detection in Heterogeneous Remote Sensing Images | The full version of this work is submitted to IEEE TGRS | IEEE Transactions on Geoscience and Remote Sensing, vol. 63, pp.
1-17, 2025, Art no. 4700817 | 10.1109/TGRS.2024.3524639 | null | cs.CV cs.LG eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Change detection (CD) in heterogeneous remote sensing images has been widely
used for disaster monitoring and land-use management. In the past decade, the
heterogeneous CD problem has significantly benefited from the development of
deep neural networks (DNNs). However, the purely data-driven DNNs perform like
a black box where the lack of interpretability limits the trustworthiness and
controllability of DNNs in most practical CD applications. As a powerful
knowledge-driven tool, copula theory performs well in modeling relationships
among random variables. To enhance the interpretability of existing neural
networks for CD, we propose a knowledge-data-driven heterogeneous CD method
based on a copula-guided neural network, named NN-Copula-CD. In our
NN-Copula-CD, the mathematical characteristics of copula are employed as the
loss functions to supervise a neural network to learn the dependence between
bi-temporal heterogeneous superpixel pairs, and then the changed regions are
identified via binary classification based on the degrees of dependence of all
the superpixel pairs in the bi-temporal images. We conduct in-depth experiments
on three datasets with heterogeneous images, where both quantitative and visual
results demonstrate the effectiveness of our proposed NN-Copula-CD method.
| [
{
"version": "v1",
"created": "Thu, 30 Mar 2023 15:20:21 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Sep 2024 14:51:30 GMT"
},
{
"version": "v3",
"created": "Thu, 19 Sep 2024 14:32:45 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Li",
"Weiming",
""
],
[
"Wang",
"Xueqian",
""
],
[
"Li",
"Gang",
""
],
[
"Geng",
"Baocheng",
""
],
[
"Varshney",
"Pramod K.",
""
]
] | TITLE: NN-Copula-CD: A Copula-Guided Interpretable Neural Network for Change
Detection in Heterogeneous Remote Sensing Images
ABSTRACT: Change detection (CD) in heterogeneous remote sensing images has been widely
used for disaster monitoring and land-use management. In the past decade, the
heterogeneous CD problem has significantly benefited from the development of
deep neural networks (DNNs). However, the purely data-driven DNNs perform like
a black box where the lack of interpretability limits the trustworthiness and
controllability of DNNs in most practical CD applications. As a powerful
knowledge-driven tool, copula theory performs well in modeling relationships
among random variables. To enhance the interpretability of existing neural
networks for CD, we propose a knowledge-data-driven heterogeneous CD method
based on a copula-guided neural network, named NN-Copula-CD. In our
NN-Copula-CD, the mathematical characteristics of copula are employed as the
loss functions to supervise a neural network to learn the dependence between
bi-temporal heterogeneous superpixel pairs, and then the changed regions are
identified via binary classification based on the degrees of dependence of all
the superpixel pairs in the bi-temporal images. We conduct in-depth experiments
on three datasets with heterogeneous images, where both quantitative and visual
results demonstrate the effectiveness of our proposed NN-Copula-CD method.
|
2305.17063 | Felix Jimenez | Felix Jimenez, Matthias Katzfuss | Vecchia Gaussian Process Ensembles on Internal Representations of Deep
Neural Networks | 22 pages, 9 figures | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | For regression tasks, standard Gaussian processes (GPs) provide natural
uncertainty quantification (UQ), while deep neural networks (DNNs) excel at
representation learning. Deterministic UQ methods for neural networks have
successfully combined the two and require only a single pass through the neural
network. However, current methods necessitate changes to network training to
address feature collapse, where unique inputs map to identical feature vectors.
We propose an alternative solution, the deep Vecchia ensemble (DVE), which
allows deterministic UQ to work in the presence of feature collapse, negating
the need for network retraining. DVE comprises an ensemble of GPs built on
hidden-layer outputs of a DNN, achieving scalability via Vecchia approximations
that leverage nearest-neighbor conditional independence. DVE is compatible with
pretrained networks and incurs low computational overhead. We demonstrate DVE's
utility on several datasets and carry out experiments to understand the inner
workings of the proposed method.
| [
{
"version": "v1",
"created": "Fri, 26 May 2023 16:19:26 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 16:50:47 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Jimenez",
"Felix",
""
],
[
"Katzfuss",
"Matthias",
""
]
] | TITLE: Vecchia Gaussian Process Ensembles on Internal Representations of Deep
Neural Networks
ABSTRACT: For regression tasks, standard Gaussian processes (GPs) provide natural
uncertainty quantification (UQ), while deep neural networks (DNNs) excel at
representation learning. Deterministic UQ methods for neural networks have
successfully combined the two and require only a single pass through the neural
network. However, current methods necessitate changes to network training to
address feature collapse, where unique inputs map to identical feature vectors.
We propose an alternative solution, the deep Vecchia ensemble (DVE), which
allows deterministic UQ to work in the presence of feature collapse, negating
the need for network retraining. DVE comprises an ensemble of GPs built on
hidden-layer outputs of a DNN, achieving scalability via Vecchia approximations
that leverage nearest-neighbor conditional independence. DVE is compatible with
pretrained networks and incurs low computational overhead. We demonstrate DVE's
utility on several datasets and carry out experiments to understand the inner
workings of the proposed method.
|
2307.09210 | Nathaniel Josephs | Nathaniel Josephs, Arash A. Amini, Marina Paez, and Lizhen Lin | Nested stochastic block model for simultaneously clustering networks and
nodes | null | null | null | null | stat.ME cs.SI stat.ML | http://creativecommons.org/licenses/by/4.0/ | We introduce the nested stochastic block model (NSBM) to cluster a collection
of networks while simultaneously detecting communities within each network.
NSBM has several appealing features including the ability to work on unlabeled
networks with potentially different node sets, the flexibility to model
heterogeneous communities, and the means to automatically select the number of
classes for the networks and the number of communities within each network.
This is accomplished via a Bayesian model, with a novel application of the
nested Dirichlet process (NDP) as a prior to jointly model the between-network
and within-network clusters. The dependency introduced by the network data
creates nontrivial challenges for the NDP, especially in the development of
efficient samplers. For posterior inference, we propose several Markov chain
Monte Carlo algorithms including a standard Gibbs sampler, a collapsed Gibbs
sampler, and two blocked Gibbs samplers that ultimately return two levels of
clustering labels from both within and across the networks. Extensive
simulation studies are carried out which demonstrate that the model provides
very accurate estimates of both levels of the clustering structure. We also
apply our model to two social network datasets that cannot be analyzed using
any previous method in the literature due to the anonymity of the nodes and the
varying number of nodes in each network.
| [
{
"version": "v1",
"created": "Tue, 18 Jul 2023 12:46:34 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 13:40:37 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Josephs",
"Nathaniel",
""
],
[
"Amini",
"Arash A.",
""
],
[
"Paez",
"Marina",
""
],
[
"Lin",
"Lizhen",
""
]
] | TITLE: Nested stochastic block model for simultaneously clustering networks and
nodes
ABSTRACT: We introduce the nested stochastic block model (NSBM) to cluster a collection
of networks while simultaneously detecting communities within each network.
NSBM has several appealing features including the ability to work on unlabeled
networks with potentially different node sets, the flexibility to model
heterogeneous communities, and the means to automatically select the number of
classes for the networks and the number of communities within each network.
This is accomplished via a Bayesian model, with a novel application of the
nested Dirichlet process (NDP) as a prior to jointly model the between-network
and within-network clusters. The dependency introduced by the network data
creates nontrivial challenges for the NDP, especially in the development of
efficient samplers. For posterior inference, we propose several Markov chain
Monte Carlo algorithms including a standard Gibbs sampler, a collapsed Gibbs
sampler, and two blocked Gibbs samplers that ultimately return two levels of
clustering labels from both within and across the networks. Extensive
simulation studies are carried out which demonstrate that the model provides
very accurate estimates of both levels of the clustering structure. We also
apply our model to two social network datasets that cannot be analyzed using
any previous method in the literature due to the anonymity of the nodes and the
varying number of nodes in each network.
|
2308.07462 | Javier Conde | Pedro Reviriego, Javier Conde, Elena Merino-G\'omez, Gonzalo
Mart\'inez, Jos\'e Alberto Hern\'andez | Playing with words: Comparing the vocabulary and lexical diversity of
ChatGPT and humans | null | Machine Learning with Applications Volume 18, December 2024,
100602 | 10.1016/j.mlwa.2024.100602 | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The introduction of Artificial Intelligence (AI) generative language models
such as GPT (Generative Pre-trained Transformer) and tools such as ChatGPT has
triggered a revolution that can transform how text is generated. This has many
implications, for example, as AI-generated text becomes a significant fraction
of the text, would this have an effect on the language capabilities of readers
and also on the training of newer AI tools? Would it affect the evolution of
languages? Focusing on one specific aspect of the language: words; will the use
of tools such as ChatGPT increase or reduce the vocabulary used or the lexical
richness? This has implications for words, as those not included in
AI-generated content will tend to be less and less popular and may eventually
be lost. In this work, we perform an initial comparison of the vocabulary and
lexical richness of ChatGPT and humans when performing the same tasks. In more
detail, two datasets containing the answers to different types of questions
answered by ChatGPT and humans, and a third dataset in which ChatGPT
paraphrases sentences and questions are used. The analysis shows that ChatGPT
tends to use fewer distinct words and lower lexical richness than humans. These
results are very preliminary and additional datasets and ChatGPT configurations
have to be evaluated to extract more general conclusions. Therefore, further
research is needed to understand how the use of ChatGPT and more broadly
generative AI tools will affect the vocabulary and lexical richness in
different types of text and languages.
| [
{
"version": "v1",
"created": "Mon, 14 Aug 2023 21:19:44 GMT"
},
{
"version": "v2",
"created": "Thu, 31 Aug 2023 11:09:16 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 16:19:46 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Reviriego",
"Pedro",
""
],
[
"Conde",
"Javier",
""
],
[
"Merino-Gómez",
"Elena",
""
],
[
"Martínez",
"Gonzalo",
""
],
[
"Hernández",
"José Alberto",
""
]
] | TITLE: Playing with words: Comparing the vocabulary and lexical diversity of
ChatGPT and humans
ABSTRACT: The introduction of Artificial Intelligence (AI) generative language models
such as GPT (Generative Pre-trained Transformer) and tools such as ChatGPT has
triggered a revolution that can transform how text is generated. This has many
implications, for example, as AI-generated text becomes a significant fraction
of the text, would this have an effect on the language capabilities of readers
and also on the training of newer AI tools? Would it affect the evolution of
languages? Focusing on one specific aspect of the language: words; will the use
of tools such as ChatGPT increase or reduce the vocabulary used or the lexical
richness? This has implications for words, as those not included in
AI-generated content will tend to be less and less popular and may eventually
be lost. In this work, we perform an initial comparison of the vocabulary and
lexical richness of ChatGPT and humans when performing the same tasks. In more
detail, two datasets containing the answers to different types of questions
answered by ChatGPT and humans, and a third dataset in which ChatGPT
paraphrases sentences and questions are used. The analysis shows that ChatGPT
tends to use fewer distinct words and lower lexical richness than humans. These
results are very preliminary and additional datasets and ChatGPT configurations
have to be evaluated to extract more general conclusions. Therefore, further
research is needed to understand how the use of ChatGPT and more broadly
generative AI tools will affect the vocabulary and lexical richness in
different types of text and languages.
|
2309.07068 | Tongkun Liu | Tongkun Liu, Bing Li, Xiao Du, Bingke Jiang, Leqi Geng, Feiyang Wang,
Zhuo Zhao | FAIR: Frequency-aware Image Restoration for Industrial Visual Anomaly
Detection | 12 pages, 10 figures | null | 10.1016/j.aei.2024.103064 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image reconstruction-based anomaly detection models are widely explored in
industrial visual inspection. However, existing models usually suffer from the
trade-off between normal reconstruction fidelity and abnormal reconstruction
distinguishability, which damages the performance. In this paper, we find that
the above trade-off can be better mitigated by leveraging the distinct
frequency biases between normal and abnormal reconstruction errors. To this
end, we propose Frequency-aware Image Restoration (FAIR), a novel
self-supervised image restoration task that restores images from their
high-frequency components. It enables precise reconstruction of normal patterns
while mitigating unfavorable generalization to anomalies. Using only a simple
vanilla UNet, FAIR achieves state-of-the-art performance with higher efficiency
on various defect detection datasets. Code: https://github.com/liutongkun/FAIR.
| [
{
"version": "v1",
"created": "Wed, 13 Sep 2023 16:28:43 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Liu",
"Tongkun",
""
],
[
"Li",
"Bing",
""
],
[
"Du",
"Xiao",
""
],
[
"Jiang",
"Bingke",
""
],
[
"Geng",
"Leqi",
""
],
[
"Wang",
"Feiyang",
""
],
[
"Zhao",
"Zhuo",
""
]
] | TITLE: FAIR: Frequency-aware Image Restoration for Industrial Visual Anomaly
Detection
ABSTRACT: Image reconstruction-based anomaly detection models are widely explored in
industrial visual inspection. However, existing models usually suffer from the
trade-off between normal reconstruction fidelity and abnormal reconstruction
distinguishability, which damages the performance. In this paper, we find that
the above trade-off can be better mitigated by leveraging the distinct
frequency biases between normal and abnormal reconstruction errors. To this
end, we propose Frequency-aware Image Restoration (FAIR), a novel
self-supervised image restoration task that restores images from their
high-frequency components. It enables precise reconstruction of normal patterns
while mitigating unfavorable generalization to anomalies. Using only a simple
vanilla UNet, FAIR achieves state-of-the-art performance with higher efficiency
on various defect detection datasets. Code: https://github.com/liutongkun/FAIR.
|
2310.01035 | Hu Wang | Hu Wang, Congbo Ma, Jianpeng Zhang, Yuan Zhang, Jodie Avery, Louise
Hull, Gustavo Carneiro | Learnable Cross-modal Knowledge Distillation for Multi-modal Learning
with Missing Modality | null | Medical Image Computing and Computer-Assisted Intervention 2023
(MICCAI 2023) | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The problem of missing modalities is both critical and non-trivial to be
handled in multi-modal models. It is common for multi-modal tasks that certain
modalities contribute more compared to other modalities, and if those important
modalities are missing, the model performance drops significantly. Such fact
remains unexplored by current multi-modal approaches that recover the
representation from missing modalities by feature reconstruction or blind
feature aggregation from other modalities, instead of extracting useful
information from the best performing modalities. In this paper, we propose a
Learnable Cross-modal Knowledge Distillation (LCKD) model to adaptively
identify important modalities and distil knowledge from them to help other
modalities from the cross-modal perspective for solving the missing modality
issue. Our approach introduces a teacher election procedure to select the most
``qualified'' teachers based on their single modality performance on certain
tasks. Then, cross-modal knowledge distillation is performed between teacher
and student modalities for each task to push the model parameters to a point
that is beneficial for all tasks. Hence, even if the teacher modalities for
certain tasks are missing during testing, the available student modalities can
accomplish the task well enough based on the learned knowledge from their
automatically elected teacher modalities. Experiments on the Brain Tumour
Segmentation Dataset 2018 (BraTS2018) shows that LCKD outperforms other methods
by a considerable margin, improving the state-of-the-art performance by 3.61%
for enhancing tumour, 5.99% for tumour core, and 3.76% for whole tumour in
terms of segmentation Dice score.
| [
{
"version": "v1",
"created": "Mon, 2 Oct 2023 09:24:54 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 07:53:09 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Wang",
"Hu",
""
],
[
"Ma",
"Congbo",
""
],
[
"Zhang",
"Jianpeng",
""
],
[
"Zhang",
"Yuan",
""
],
[
"Avery",
"Jodie",
""
],
[
"Hull",
"Louise",
""
],
[
"Carneiro",
"Gustavo",
""
]
] | TITLE: Learnable Cross-modal Knowledge Distillation for Multi-modal Learning
with Missing Modality
ABSTRACT: The problem of missing modalities is both critical and non-trivial to be
handled in multi-modal models. It is common for multi-modal tasks that certain
modalities contribute more compared to other modalities, and if those important
modalities are missing, the model performance drops significantly. Such fact
remains unexplored by current multi-modal approaches that recover the
representation from missing modalities by feature reconstruction or blind
feature aggregation from other modalities, instead of extracting useful
information from the best performing modalities. In this paper, we propose a
Learnable Cross-modal Knowledge Distillation (LCKD) model to adaptively
identify important modalities and distil knowledge from them to help other
modalities from the cross-modal perspective for solving the missing modality
issue. Our approach introduces a teacher election procedure to select the most
``qualified'' teachers based on their single modality performance on certain
tasks. Then, cross-modal knowledge distillation is performed between teacher
and student modalities for each task to push the model parameters to a point
that is beneficial for all tasks. Hence, even if the teacher modalities for
certain tasks are missing during testing, the available student modalities can
accomplish the task well enough based on the learned knowledge from their
automatically elected teacher modalities. Experiments on the Brain Tumour
Segmentation Dataset 2018 (BraTS2018) shows that LCKD outperforms other methods
by a considerable margin, improving the state-of-the-art performance by 3.61%
for enhancing tumour, 5.99% for tumour core, and 3.76% for whole tumour in
terms of segmentation Dice score.
|
2310.14560 | Hui Tian | Hui Tian, Kai Xu | Polyhedral Surface: Self-supervised Point Cloud Reconstruction Based on
Polyhedral Surface | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Point cloud reconstruction from raw point cloud has been an important topic
in computer graphics for decades, especially due to its high demand in modeling
and rendering applications. An important way to solve this problem is
establishing a local geometry to fit the local curve. However, previous methods
build either a local plane or polynomial curve. Local plane brings the loss of
sharp feature and the boundary artefacts on open surface. Polynomial curve is
hard to combine with neural network due to the local coordinate consistent
problem. To address this, we propose a novel polyhedral surface to represent
local surface. This method provides more flexible to represent sharp feature
and surface boundary on open surface. It does not require any local coordinate
system, which is important when introducing neural networks. Specifically, we
use normals to construct the polyhedral surface, including both dihedral and
trihedral surfaces using 2 and 3 normals, respectively. Our method achieves
state-of-the-art results on three commonly used datasets (ShapeNetCore, ABC,
and ScanNet). Code will be released upon acceptance.
| [
{
"version": "v1",
"created": "Mon, 23 Oct 2023 04:24:31 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 12:25:32 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Tian",
"Hui",
""
],
[
"Xu",
"Kai",
""
]
] | TITLE: Polyhedral Surface: Self-supervised Point Cloud Reconstruction Based on
Polyhedral Surface
ABSTRACT: Point cloud reconstruction from raw point cloud has been an important topic
in computer graphics for decades, especially due to its high demand in modeling
and rendering applications. An important way to solve this problem is
establishing a local geometry to fit the local curve. However, previous methods
build either a local plane or polynomial curve. Local plane brings the loss of
sharp feature and the boundary artefacts on open surface. Polynomial curve is
hard to combine with neural network due to the local coordinate consistent
problem. To address this, we propose a novel polyhedral surface to represent
local surface. This method provides more flexible to represent sharp feature
and surface boundary on open surface. It does not require any local coordinate
system, which is important when introducing neural networks. Specifically, we
use normals to construct the polyhedral surface, including both dihedral and
trihedral surfaces using 2 and 3 normals, respectively. Our method achieves
state-of-the-art results on three commonly used datasets (ShapeNetCore, ABC,
and ScanNet). Code will be released upon acceptance.
|
2310.20335 | Gonzalo Contreras-Aso | Gonzalo Contreras-Aso, Cristian P\'erez-Corral, Miguel Romance | Uplifting edges in higher order networks: spectral centralities for
non-uniform hypergraphs | 29 pages, 8 figures | null | 10.3934/math.20241539 | null | math.SP math-ph math.MP physics.comp-ph physics.soc-ph | http://creativecommons.org/licenses/by/4.0/ | Spectral analysis of networks states that many structural properties of
graphs, such as centrality of their nodes, are given in terms of their
adjacency matrices. The natural extension of such spectral analysis to higher
order networks is strongly limited by the fact that a given hypergraph could
have several different adjacency hypermatrices, hence the results obtained so
far are mainly restricted to the class of uniform hypergraphs, which leaves
many real systems unattended. A new method for analysing non-linear
eigenvector-like centrality measures of non-uniform hypergraphs is presented in
this paper that could be useful for studying properties of
$\mathcal{H}$-eigenvectors and $\mathcal{Z}$-eigenvectors in the non-uniform
case. In order to do so, a new operation - the $\textit{uplift}$ - is
introduced, incorporating auxiliary nodes in the hypergraph to allow for a
uniform-like analysis. We later argue why this is a mathematically sound
operation, and we furthermore use it to classify a whole family of hypergraphs
with unique Perron-like $\mathcal{Z}$-eigenvectors. We supplement the
theoretical analysis with several examples and numerical simulations on
synthetic and real datasets.
| [
{
"version": "v1",
"created": "Tue, 31 Oct 2023 10:21:58 GMT"
},
{
"version": "v2",
"created": "Mon, 15 Jul 2024 10:15:46 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Contreras-Aso",
"Gonzalo",
""
],
[
"Pérez-Corral",
"Cristian",
""
],
[
"Romance",
"Miguel",
""
]
] | TITLE: Uplifting edges in higher order networks: spectral centralities for
non-uniform hypergraphs
ABSTRACT: Spectral analysis of networks states that many structural properties of
graphs, such as centrality of their nodes, are given in terms of their
adjacency matrices. The natural extension of such spectral analysis to higher
order networks is strongly limited by the fact that a given hypergraph could
have several different adjacency hypermatrices, hence the results obtained so
far are mainly restricted to the class of uniform hypergraphs, which leaves
many real systems unattended. A new method for analysing non-linear
eigenvector-like centrality measures of non-uniform hypergraphs is presented in
this paper that could be useful for studying properties of
$\mathcal{H}$-eigenvectors and $\mathcal{Z}$-eigenvectors in the non-uniform
case. In order to do so, a new operation - the $\textit{uplift}$ - is
introduced, incorporating auxiliary nodes in the hypergraph to allow for a
uniform-like analysis. We later argue why this is a mathematically sound
operation, and we furthermore use it to classify a whole family of hypergraphs
with unique Perron-like $\mathcal{Z}$-eigenvectors. We supplement the
theoretical analysis with several examples and numerical simulations on
synthetic and real datasets.
|
2311.05029 | Christian Wilms | Robert Johanson and Christian Wilms and Ole Johannsen and Simone
Frintrop | S$^3$AD: Semi-supervised Small Apple Detection in Orchard Environments | Accepted at WACV 2024. The code and the dataset MAD are available at
http://www.inf.uni-hamburg.de/mad | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Crop detection is integral for precision agriculture applications such as
automated yield estimation or fruit picking. However, crop detection, e.g.,
apple detection in orchard environments remains challenging due to a lack of
large-scale datasets and the small relative size of the crops in the image. In
this work, we address these challenges by reformulating the apple detection
task in a semi-supervised manner. To this end, we provide the large,
high-resolution dataset MAD comprising 105 labeled images with 14,667 annotated
apple instances and 4,440 unlabeled images. Utilizing this dataset, we also
propose a novel Semi-Supervised Small Apple Detection system S$^3$AD based on
contextual attention and selective tiling to improve the challenging detection
of small apples, while limiting the computational overhead. We conduct an
extensive evaluation on MAD and the MSU dataset, showing that S$^3$AD
substantially outperforms strong fully-supervised baselines, including several
small object detection systems, by up to $14.9\%$. Additionally, we exploit the
detailed annotations of our dataset w.r.t. apple properties to analyze the
influence of relative size or level of occlusion on the results of various
systems, quantifying current challenges.
| [
{
"version": "v1",
"created": "Wed, 8 Nov 2023 21:25:27 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 09:10:28 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Johanson",
"Robert",
""
],
[
"Wilms",
"Christian",
""
],
[
"Johannsen",
"Ole",
""
],
[
"Frintrop",
"Simone",
""
]
] | TITLE: S$^3$AD: Semi-supervised Small Apple Detection in Orchard Environments
ABSTRACT: Crop detection is integral for precision agriculture applications such as
automated yield estimation or fruit picking. However, crop detection, e.g.,
apple detection in orchard environments remains challenging due to a lack of
large-scale datasets and the small relative size of the crops in the image. In
this work, we address these challenges by reformulating the apple detection
task in a semi-supervised manner. To this end, we provide the large,
high-resolution dataset MAD comprising 105 labeled images with 14,667 annotated
apple instances and 4,440 unlabeled images. Utilizing this dataset, we also
propose a novel Semi-Supervised Small Apple Detection system S$^3$AD based on
contextual attention and selective tiling to improve the challenging detection
of small apples, while limiting the computational overhead. We conduct an
extensive evaluation on MAD and the MSU dataset, showing that S$^3$AD
substantially outperforms strong fully-supervised baselines, including several
small object detection systems, by up to $14.9\%$. Additionally, we exploit the
detailed annotations of our dataset w.r.t. apple properties to analyze the
influence of relative size or level of occlusion on the results of various
systems, quantifying current challenges.
|
2312.04584 | Yiming Li | Mingyan Zhu, Yiming Li, Junfeng Guo, Tao Wei, Shu-Tao Xia, Zhan Qin | Towards Sample-specific Backdoor Attack with Clean Labels via Attribute
Trigger | This paper is accepted by IEEE Transactions on Dependable and Secure
Computing (TDSC), 2025. The first two authors contributed equally to this
work. 14 pages | null | null | null | cs.CR cs.AI cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Currently, sample-specific backdoor attacks (SSBAs) are the most advanced and
malicious methods since they can easily circumvent most of the current backdoor
defenses. In this paper, we reveal that SSBAs are not sufficiently stealthy due
to their poisoned-label nature, where users can discover anomalies if they
check the image-label relationship. In particular, we demonstrate that it is
ineffective to directly generalize existing SSBAs to their clean-label variants
by poisoning samples solely from the target class. We reveal that it is
primarily due to two reasons, including \textbf{(1)} the `antagonistic effects'
of ground-truth features and \textbf{(2)} the learning difficulty of
sample-specific features. Accordingly, trigger-related features of existing
SSBAs cannot be effectively learned under the clean-label setting due to their
mild trigger intensity required for ensuring stealthiness. We argue that the
intensity constraint of existing SSBAs is mostly because their trigger patterns
are `content-irrelevant' and therefore act as `noises' for both humans and
DNNs. Motivated by this understanding, we propose to exploit content-relevant
features, $a.k.a.$ (human-relied) attributes, as the trigger patterns to design
clean-label SSBAs. This new attack paradigm is dubbed backdoor attack with
attribute trigger (BAAT). Extensive experiments are conducted on benchmark
datasets, which verify the effectiveness of our BAAT and its resistance to
existing defenses.
| [
{
"version": "v1",
"created": "Sun, 3 Dec 2023 09:12:14 GMT"
},
{
"version": "v2",
"created": "Mon, 11 Dec 2023 03:12:36 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 13:36:51 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Zhu",
"Mingyan",
""
],
[
"Li",
"Yiming",
""
],
[
"Guo",
"Junfeng",
""
],
[
"Wei",
"Tao",
""
],
[
"Xia",
"Shu-Tao",
""
],
[
"Qin",
"Zhan",
""
]
] | TITLE: Towards Sample-specific Backdoor Attack with Clean Labels via Attribute
Trigger
ABSTRACT: Currently, sample-specific backdoor attacks (SSBAs) are the most advanced and
malicious methods since they can easily circumvent most of the current backdoor
defenses. In this paper, we reveal that SSBAs are not sufficiently stealthy due
to their poisoned-label nature, where users can discover anomalies if they
check the image-label relationship. In particular, we demonstrate that it is
ineffective to directly generalize existing SSBAs to their clean-label variants
by poisoning samples solely from the target class. We reveal that it is
primarily due to two reasons, including \textbf{(1)} the `antagonistic effects'
of ground-truth features and \textbf{(2)} the learning difficulty of
sample-specific features. Accordingly, trigger-related features of existing
SSBAs cannot be effectively learned under the clean-label setting due to their
mild trigger intensity required for ensuring stealthiness. We argue that the
intensity constraint of existing SSBAs is mostly because their trigger patterns
are `content-irrelevant' and therefore act as `noises' for both humans and
DNNs. Motivated by this understanding, we propose to exploit content-relevant
features, $a.k.a.$ (human-relied) attributes, as the trigger patterns to design
clean-label SSBAs. This new attack paradigm is dubbed backdoor attack with
attribute trigger (BAAT). Extensive experiments are conducted on benchmark
datasets, which verify the effectiveness of our BAAT and its resistance to
existing defenses.
|
2312.05327 | Ribana Roscher | Ribana Roscher and Marc Ru{\ss}wurm and Caroline Gevaert and Michael
Kampffmeyer and Jefersson A. dos Santos and Maria Vakalopoulou and Ronny
H\"ansch and Stine Hansen and Keiller Nogueira and Jonathan Prexl and Devis
Tuia | Better, Not Just More: Data-Centric Machine Learning for Earth
Observation | null | IEEE Geoscience and Remote Sensing Magazine, vol. 12, no. 4, pp.
335-355, Dec. 2024 | 10.1109/MGRS.2024.3470986 | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent developments and research in modern machine learning have led to
substantial improvements in the geospatial field. Although numerous deep
learning architectures and models have been proposed, the majority of them have
been solely developed on benchmark datasets that lack strong real-world
relevance. Furthermore, the performance of many methods has already saturated
on these datasets. We argue that a shift from a model-centric view to a
complementary data-centric perspective is necessary for further improvements in
accuracy, generalization ability, and real impact on end-user applications.
Furthermore, considering the entire machine learning cycle-from problem
definition to model deployment with feedback-is crucial for enhancing machine
learning models that can be reliable in unforeseen situations. This work
presents a definition as well as a precise categorization and overview of
automated data-centric learning approaches for geospatial data. It highlights
the complementary role of data-centric learning with respect to model-centric
in the larger machine learning deployment cycle. We review papers across the
entire geospatial field and categorize them into different groups. A set of
representative experiments shows concrete implementation examples. These
examples provide concrete steps to act on geospatial data with data-centric
machine learning approaches.
| [
{
"version": "v1",
"created": "Fri, 8 Dec 2023 19:24:05 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Jun 2024 16:29:56 GMT"
},
{
"version": "v3",
"created": "Tue, 5 Nov 2024 14:12:40 GMT"
},
{
"version": "v4",
"created": "Thu, 13 Mar 2025 20:09:55 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Roscher",
"Ribana",
""
],
[
"Rußwurm",
"Marc",
""
],
[
"Gevaert",
"Caroline",
""
],
[
"Kampffmeyer",
"Michael",
""
],
[
"Santos",
"Jefersson A. dos",
""
],
[
"Vakalopoulou",
"Maria",
""
],
[
"Hänsch",
"Ronny",
""
],
[
"Hansen",
"Stine",
""
],
[
"Nogueira",
"Keiller",
""
],
[
"Prexl",
"Jonathan",
""
],
[
"Tuia",
"Devis",
""
]
] | TITLE: Better, Not Just More: Data-Centric Machine Learning for Earth
Observation
ABSTRACT: Recent developments and research in modern machine learning have led to
substantial improvements in the geospatial field. Although numerous deep
learning architectures and models have been proposed, the majority of them have
been solely developed on benchmark datasets that lack strong real-world
relevance. Furthermore, the performance of many methods has already saturated
on these datasets. We argue that a shift from a model-centric view to a
complementary data-centric perspective is necessary for further improvements in
accuracy, generalization ability, and real impact on end-user applications.
Furthermore, considering the entire machine learning cycle-from problem
definition to model deployment with feedback-is crucial for enhancing machine
learning models that can be reliable in unforeseen situations. This work
presents a definition as well as a precise categorization and overview of
automated data-centric learning approaches for geospatial data. It highlights
the complementary role of data-centric learning with respect to model-centric
in the larger machine learning deployment cycle. We review papers across the
entire geospatial field and categorize them into different groups. A set of
representative experiments shows concrete implementation examples. These
examples provide concrete steps to act on geospatial data with data-centric
machine learning approaches.
|
2312.15268 | Kaichen Zhou | Kaichen Zhou, Jia-Wang Bian, Jian-Qing Zheng, Jiaxing Zhong, Qian Xie,
Niki Trigoni, Andrew Markham | Manydepth2: Motion-Aware Self-Supervised Monocular Depth Estimation in
Dynamic Scenes | Monocular Depth Estimation, Self-Supervised, Optical Flow | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite advancements in self-supervised monocular depth estimation,
challenges persist in dynamic scenarios due to the dependence on assumptions
about a static world. In this paper, we present Manydepth2, to achieve precise
depth estimation for both dynamic objects and static backgrounds, all while
maintaining computational efficiency. To tackle the challenges posed by dynamic
content, we incorporate optical flow and coarse monocular depth to create a
pseudo-static reference frame. This frame is then utilized to build a
motion-aware cost volume in collaboration with the vanilla target frame.
Furthermore, to improve the accuracy and robustness of the network
architecture, we propose an attention-based depth network that effectively
integrates information from feature maps at different resolutions by
incorporating both channel and non-local attention mechanisms. Compared to
methods with similar computational costs, Manydepth2 achieves a significant
reduction of approximately five percent in root-mean-square error for
self-supervised monocular depth estimation on the KITTI-2015 dataset. The code
could be found at https://github.com/kaichen-z/Manydepth2.
| [
{
"version": "v1",
"created": "Sat, 23 Dec 2023 14:36:27 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Sep 2024 17:45:13 GMT"
},
{
"version": "v3",
"created": "Sat, 21 Sep 2024 17:46:31 GMT"
},
{
"version": "v4",
"created": "Thu, 26 Sep 2024 15:10:58 GMT"
},
{
"version": "v5",
"created": "Sun, 29 Sep 2024 01:16:33 GMT"
},
{
"version": "v6",
"created": "Fri, 11 Oct 2024 23:24:26 GMT"
},
{
"version": "v7",
"created": "Sat, 18 Jan 2025 16:16:10 GMT"
},
{
"version": "v8",
"created": "Tue, 28 Jan 2025 01:18:32 GMT"
},
{
"version": "v9",
"created": "Thu, 13 Mar 2025 19:28:00 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Zhou",
"Kaichen",
""
],
[
"Bian",
"Jia-Wang",
""
],
[
"Zheng",
"Jian-Qing",
""
],
[
"Zhong",
"Jiaxing",
""
],
[
"Xie",
"Qian",
""
],
[
"Trigoni",
"Niki",
""
],
[
"Markham",
"Andrew",
""
]
] | TITLE: Manydepth2: Motion-Aware Self-Supervised Monocular Depth Estimation in
Dynamic Scenes
ABSTRACT: Despite advancements in self-supervised monocular depth estimation,
challenges persist in dynamic scenarios due to the dependence on assumptions
about a static world. In this paper, we present Manydepth2, to achieve precise
depth estimation for both dynamic objects and static backgrounds, all while
maintaining computational efficiency. To tackle the challenges posed by dynamic
content, we incorporate optical flow and coarse monocular depth to create a
pseudo-static reference frame. This frame is then utilized to build a
motion-aware cost volume in collaboration with the vanilla target frame.
Furthermore, to improve the accuracy and robustness of the network
architecture, we propose an attention-based depth network that effectively
integrates information from feature maps at different resolutions by
incorporating both channel and non-local attention mechanisms. Compared to
methods with similar computational costs, Manydepth2 achieves a significant
reduction of approximately five percent in root-mean-square error for
self-supervised monocular depth estimation on the KITTI-2015 dataset. The code
could be found at https://github.com/kaichen-z/Manydepth2.
|
2402.07594 | Ramses Sanchez | Patrick Seifner, Kostadin Cvejoski, Antonia K\"orner, Rams\'es J.
S\'anchez | Zero-shot Imputation with Foundation Inference Models for Dynamical
Systems | null | null | null | null | cs.LG math.DS | http://creativecommons.org/licenses/by/4.0/ | Dynamical systems governed by ordinary differential equations (ODEs) serve as
models for a vast number of natural and social phenomena. In this work, we
offer a fresh perspective on the classical problem of imputing missing time
series data, whose underlying dynamics are assumed to be determined by ODEs.
Specifically, we revisit ideas from amortized inference and neural operators,
and propose a novel supervised learning framework for zero-shot time series
imputation, through parametric functions satisfying some (hidden) ODEs. Our
proposal consists of two components. First, a broad probability distribution
over the space of ODE solutions, observation times and noise mechanisms, with
which we generate a large, synthetic dataset of (hidden) ODE solutions, along
with their noisy and sparse observations. Second, a neural recognition model
that is trained offline, to map the generated time series onto the spaces of
initial conditions and time derivatives of the (hidden) ODE solutions, which we
then integrate to impute the missing data. We empirically demonstrate that one
and the same (pretrained) recognition model can perform zero-shot imputation
across 63 distinct time series with missing values, each sampled from widely
different dynamical systems. Likewise, we demonstrate that it can perform
zero-shot imputation of missing high-dimensional data in 10 vastly different
settings, spanning human motion, air quality, traffic and electricity studies,
as well as Navier-Stokes simulations -- without requiring any fine-tuning. What
is more, our proposal often outperforms state-of-the-art methods, which are
trained on the target datasets.
Our pretrained model, repository and tutorials are available online.
| [
{
"version": "v1",
"created": "Mon, 12 Feb 2024 11:48:54 GMT"
},
{
"version": "v2",
"created": "Fri, 4 Oct 2024 10:41:18 GMT"
},
{
"version": "v3",
"created": "Fri, 28 Feb 2025 14:24:39 GMT"
},
{
"version": "v4",
"created": "Fri, 14 Mar 2025 15:37:14 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Seifner",
"Patrick",
""
],
[
"Cvejoski",
"Kostadin",
""
],
[
"Körner",
"Antonia",
""
],
[
"Sánchez",
"Ramsés J.",
""
]
] | TITLE: Zero-shot Imputation with Foundation Inference Models for Dynamical
Systems
ABSTRACT: Dynamical systems governed by ordinary differential equations (ODEs) serve as
models for a vast number of natural and social phenomena. In this work, we
offer a fresh perspective on the classical problem of imputing missing time
series data, whose underlying dynamics are assumed to be determined by ODEs.
Specifically, we revisit ideas from amortized inference and neural operators,
and propose a novel supervised learning framework for zero-shot time series
imputation, through parametric functions satisfying some (hidden) ODEs. Our
proposal consists of two components. First, a broad probability distribution
over the space of ODE solutions, observation times and noise mechanisms, with
which we generate a large, synthetic dataset of (hidden) ODE solutions, along
with their noisy and sparse observations. Second, a neural recognition model
that is trained offline, to map the generated time series onto the spaces of
initial conditions and time derivatives of the (hidden) ODE solutions, which we
then integrate to impute the missing data. We empirically demonstrate that one
and the same (pretrained) recognition model can perform zero-shot imputation
across 63 distinct time series with missing values, each sampled from widely
different dynamical systems. Likewise, we demonstrate that it can perform
zero-shot imputation of missing high-dimensional data in 10 vastly different
settings, spanning human motion, air quality, traffic and electricity studies,
as well as Navier-Stokes simulations -- without requiring any fine-tuning. What
is more, our proposal often outperforms state-of-the-art methods, which are
trained on the target datasets.
Our pretrained model, repository and tutorials are available online.
|
2402.08635 | Hasan Mahmud | Husne Ara Rubaiyeat, Hasan Mahmud, Ahsan Habib, Md. Kamrul Hasan | BdSLW60: A Word-Level Bangla Sign Language Dataset | null | Accpeted by Multimedia Tools and Applications 2025 | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Sign language discourse is an essential mode of daily communication for the
deaf and hard-of-hearing people. However, research on Bangla Sign Language
(BdSL) faces notable limitations, primarily due to the lack of datasets.
Recognizing wordlevel signs in BdSL (WL-BdSL) presents a multitude of
challenges, including the need for well-annotated datasets, capturing the
dynamic nature of sign gestures from facial or hand landmarks, developing
suitable machine learning or deep learning-based models with substantial video
samples, and so on. In this paper, we address these challenges by creating a
comprehensive BdSL word-level dataset named BdSLW60 in an unconstrained and
natural setting, allowing positional and temporal variations and allowing sign
users to change hand dominance freely. The dataset encompasses 60 Bangla sign
words, with a significant scale of 9307 video trials provided by 18 signers
under the supervision of a sign language professional. The dataset was
rigorously annotated and cross-checked by 60 annotators. We also introduced a
unique approach of a relative quantization-based key frame encoding technique
for landmark based sign gesture recognition. We report the benchmarking of our
BdSLW60 dataset using the Support Vector Machine (SVM) with testing accuracy up
to 67.6% and an attention-based bi-LSTM with testing accuracy up to 75.1%. The
dataset is available at https://www.kaggle.com/datasets/hasaniut/bdslw60 and
the code base is accessible from https://github.com/hasanssl/BdSLW60_Code.
| [
{
"version": "v1",
"created": "Tue, 13 Feb 2024 18:02:58 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Rubaiyeat",
"Husne Ara",
""
],
[
"Mahmud",
"Hasan",
""
],
[
"Habib",
"Ahsan",
""
],
[
"Hasan",
"Md. Kamrul",
""
]
] | TITLE: BdSLW60: A Word-Level Bangla Sign Language Dataset
ABSTRACT: Sign language discourse is an essential mode of daily communication for the
deaf and hard-of-hearing people. However, research on Bangla Sign Language
(BdSL) faces notable limitations, primarily due to the lack of datasets.
Recognizing wordlevel signs in BdSL (WL-BdSL) presents a multitude of
challenges, including the need for well-annotated datasets, capturing the
dynamic nature of sign gestures from facial or hand landmarks, developing
suitable machine learning or deep learning-based models with substantial video
samples, and so on. In this paper, we address these challenges by creating a
comprehensive BdSL word-level dataset named BdSLW60 in an unconstrained and
natural setting, allowing positional and temporal variations and allowing sign
users to change hand dominance freely. The dataset encompasses 60 Bangla sign
words, with a significant scale of 9307 video trials provided by 18 signers
under the supervision of a sign language professional. The dataset was
rigorously annotated and cross-checked by 60 annotators. We also introduced a
unique approach of a relative quantization-based key frame encoding technique
for landmark based sign gesture recognition. We report the benchmarking of our
BdSLW60 dataset using the Support Vector Machine (SVM) with testing accuracy up
to 67.6% and an attention-based bi-LSTM with testing accuracy up to 75.1%. The
dataset is available at https://www.kaggle.com/datasets/hasaniut/bdslw60 and
the code base is accessible from https://github.com/hasanssl/BdSLW60_Code.
|
2403.08215 | Sicen Guo | Sicen Guo, Ziwei Long, Zhiyuan Wu, Qijun Chen, Ioannis Pitas and Rui
Fan | LIX: Implicitly Infusing Spatial Geometric Prior Knowledge into Visual
Semantic Segmentation for Autonomous Driving | 13 pages, 7 figures, 5 tables | null | null | null | cs.CV cs.AI cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the impressive performance achieved by data-fusion networks with
duplex encoders for visual semantic segmentation, they become ineffective when
spatial geometric data are not available. Implicitly infusing the spatial
geometric prior knowledge acquired by a data-fusion teacher network into a
single-modal student network is a practical, albeit less explored research
avenue. This article delves into this topic and resorts to knowledge
distillation approaches to address this problem. We introduce the Learning to
Infuse ''X'' (LIX) framework, with novel contributions in both logit
distillation and feature distillation aspects. We present a mathematical proof
that underscores the limitation of using a single, fixed weight in decoupled
knowledge distillation and introduce a logit-wise dynamic weight controller as
a solution to this issue. Furthermore, we develop an adaptively-recalibrated
feature distillation algorithm, including two novel techniques: feature
recalibration via kernel regression and in-depth feature consistency
quantification via centered kernel alignment. Extensive experiments conducted
with intermediate-fusion and late-fusion networks across various public
datasets provide both quantitative and qualitative evaluations, demonstrating
the superior performance of our LIX framework when compared to other
state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Wed, 13 Mar 2024 03:24:36 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 09:24:22 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Guo",
"Sicen",
""
],
[
"Long",
"Ziwei",
""
],
[
"Wu",
"Zhiyuan",
""
],
[
"Chen",
"Qijun",
""
],
[
"Pitas",
"Ioannis",
""
],
[
"Fan",
"Rui",
""
]
] | TITLE: LIX: Implicitly Infusing Spatial Geometric Prior Knowledge into Visual
Semantic Segmentation for Autonomous Driving
ABSTRACT: Despite the impressive performance achieved by data-fusion networks with
duplex encoders for visual semantic segmentation, they become ineffective when
spatial geometric data are not available. Implicitly infusing the spatial
geometric prior knowledge acquired by a data-fusion teacher network into a
single-modal student network is a practical, albeit less explored research
avenue. This article delves into this topic and resorts to knowledge
distillation approaches to address this problem. We introduce the Learning to
Infuse ''X'' (LIX) framework, with novel contributions in both logit
distillation and feature distillation aspects. We present a mathematical proof
that underscores the limitation of using a single, fixed weight in decoupled
knowledge distillation and introduce a logit-wise dynamic weight controller as
a solution to this issue. Furthermore, we develop an adaptively-recalibrated
feature distillation algorithm, including two novel techniques: feature
recalibration via kernel regression and in-depth feature consistency
quantification via centered kernel alignment. Extensive experiments conducted
with intermediate-fusion and late-fusion networks across various public
datasets provide both quantitative and qualitative evaluations, demonstrating
the superior performance of our LIX framework when compared to other
state-of-the-art approaches.
|
2404.02101 | Hao He | Hao He, Yinghao Xu, Yuwei Guo, Gordon Wetzstein, Bo Dai, Hongsheng Li,
Ceyuan Yang | CameraCtrl: Enabling Camera Control for Text-to-Video Generation | Project page: https://hehao13.github.io/projects-CameraCtrl/ Code:
https://github.com/hehao13/CameraCtrl | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Controllability plays a crucial role in video generation, as it allows users
to create and edit content more precisely. Existing models, however, lack
control of camera pose that serves as a cinematic language to express deeper
narrative nuances. To alleviate this issue, we introduce CameraCtrl, enabling
accurate camera pose control for video diffusion models. Our approach explores
effective camera trajectory parameterization along with a plug-and-play camera
pose control module that is trained on top of a video diffusion model, leaving
other modules of the base model untouched. Moreover, a comprehensive study on
the effect of various training datasets is conducted, suggesting that videos
with diverse camera distributions and similar appearance to the base model
indeed enhance controllability and generalization. Experimental results
demonstrate the effectiveness of CameraCtrl in achieving precise camera control
with different video generation models, marking a step forward in the pursuit
of dynamic and customized video storytelling from textual and camera pose
inputs.
| [
{
"version": "v1",
"created": "Tue, 2 Apr 2024 16:52:41 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Mar 2025 18:35:06 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"He",
"Hao",
""
],
[
"Xu",
"Yinghao",
""
],
[
"Guo",
"Yuwei",
""
],
[
"Wetzstein",
"Gordon",
""
],
[
"Dai",
"Bo",
""
],
[
"Li",
"Hongsheng",
""
],
[
"Yang",
"Ceyuan",
""
]
] | TITLE: CameraCtrl: Enabling Camera Control for Text-to-Video Generation
ABSTRACT: Controllability plays a crucial role in video generation, as it allows users
to create and edit content more precisely. Existing models, however, lack
control of camera pose that serves as a cinematic language to express deeper
narrative nuances. To alleviate this issue, we introduce CameraCtrl, enabling
accurate camera pose control for video diffusion models. Our approach explores
effective camera trajectory parameterization along with a plug-and-play camera
pose control module that is trained on top of a video diffusion model, leaving
other modules of the base model untouched. Moreover, a comprehensive study on
the effect of various training datasets is conducted, suggesting that videos
with diverse camera distributions and similar appearance to the base model
indeed enhance controllability and generalization. Experimental results
demonstrate the effectiveness of CameraCtrl in achieving precise camera control
with different video generation models, marking a step forward in the pursuit
of dynamic and customized video storytelling from textual and camera pose
inputs.
|
2405.03279 | Qizhou Chen | Qizhou Chen, Taolin Zhang, Xiaofeng He, Dongyang Li, Chengyu Wang,
Longtao Huang, Hui Xue | Lifelong Knowledge Editing for LLMs with Retrieval-Augmented Continuous
Prompt Learning | EMNLP 2024 main | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Model editing aims to correct outdated or erroneous knowledge in large
language models (LLMs) without the need for costly retraining. Lifelong model
editing is the most challenging task that caters to the continuous editing
requirements of LLMs. Prior works primarily focus on single or batch editing;
nevertheless, these methods fall short in lifelong editing scenarios due to
catastrophic knowledge forgetting and the degradation of model performance.
Although retrieval-based methods alleviate these issues, they are impeded by
slow and cumbersome processes of integrating the retrieved knowledge into the
model. In this work, we introduce RECIPE, a RetriEval-augmented ContInuous
Prompt lEarning method, to boost editing efficacy and inference efficiency in
lifelong learning. RECIPE first converts knowledge statements into short and
informative continuous prompts, prefixed to the LLM's input query embedding, to
efficiently refine the response grounded on the knowledge. It further
integrates the Knowledge Sentinel (KS) that acts as an intermediary to
calculate a dynamic threshold, determining whether the retrieval repository
contains relevant knowledge. Our retriever and prompt encoder are jointly
trained to achieve editing properties, i.e., reliability, generality, and
locality. In our experiments, RECIPE is assessed extensively across multiple
LLMs and editing datasets, where it achieves superior editing performance.
RECIPE also demonstrates its capability to maintain the overall performance of
LLMs alongside showcasing fast editing and inference speed.
| [
{
"version": "v1",
"created": "Mon, 6 May 2024 08:52:11 GMT"
},
{
"version": "v2",
"created": "Wed, 8 May 2024 03:45:51 GMT"
},
{
"version": "v3",
"created": "Fri, 4 Oct 2024 12:29:46 GMT"
},
{
"version": "v4",
"created": "Fri, 14 Mar 2025 03:56:58 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Chen",
"Qizhou",
""
],
[
"Zhang",
"Taolin",
""
],
[
"He",
"Xiaofeng",
""
],
[
"Li",
"Dongyang",
""
],
[
"Wang",
"Chengyu",
""
],
[
"Huang",
"Longtao",
""
],
[
"Xue",
"Hui",
""
]
] | TITLE: Lifelong Knowledge Editing for LLMs with Retrieval-Augmented Continuous
Prompt Learning
ABSTRACT: Model editing aims to correct outdated or erroneous knowledge in large
language models (LLMs) without the need for costly retraining. Lifelong model
editing is the most challenging task that caters to the continuous editing
requirements of LLMs. Prior works primarily focus on single or batch editing;
nevertheless, these methods fall short in lifelong editing scenarios due to
catastrophic knowledge forgetting and the degradation of model performance.
Although retrieval-based methods alleviate these issues, they are impeded by
slow and cumbersome processes of integrating the retrieved knowledge into the
model. In this work, we introduce RECIPE, a RetriEval-augmented ContInuous
Prompt lEarning method, to boost editing efficacy and inference efficiency in
lifelong learning. RECIPE first converts knowledge statements into short and
informative continuous prompts, prefixed to the LLM's input query embedding, to
efficiently refine the response grounded on the knowledge. It further
integrates the Knowledge Sentinel (KS) that acts as an intermediary to
calculate a dynamic threshold, determining whether the retrieval repository
contains relevant knowledge. Our retriever and prompt encoder are jointly
trained to achieve editing properties, i.e., reliability, generality, and
locality. In our experiments, RECIPE is assessed extensively across multiple
LLMs and editing datasets, where it achieves superior editing performance.
RECIPE also demonstrates its capability to maintain the overall performance of
LLMs alongside showcasing fast editing and inference speed.
|
2406.00009 | Hang Zhou | Hang Zhou, Ke Ma, Shixiao Liang, Xiaopeng Li, Xiaobo Qu | ULTra-AV: A Unified Longitudinal Trajectory Dataset for Automated
Vehicle | NA | null | 10.1038/s41597-024-03795-y | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated Vehicles (AVs) promise significant advances in transportation.
Critical to these improvements is understanding AVs' longitudinal behavior,
relying heavily on real-world trajectory data. Existing open-source trajectory
datasets of AV, however, often fall short in refinement, reliability, and
completeness, hindering effective performance metrics analysis and model
development. This study addresses these challenges by creating a Unified
Longitudinal TRAjectory dataset for AVs (Ultra-AV) to analyze their microscopic
longitudinal driving behaviors. This dataset compiles data from 13 distinct
sources, encompassing various AV types, test sites, and experiment scenarios.
We established a three-step data processing: 1. extraction of longitudinal
trajectory data, 2. general data cleaning, and 3. data-specific cleaning to
obtain the longitudinal trajectory data and car-following trajectory data. The
validity of the processed data is affirmed through performance evaluations
across safety, mobility, stability, and sustainability, along with an analysis
of the relationships between variables in car-following models. Our work not
only furnishes researchers with standardized data and metrics for longitudinal
AV behavior studies but also sets guidelines for data collection and model
development.
| [
{
"version": "v1",
"created": "Thu, 16 May 2024 21:03:31 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Zhou",
"Hang",
""
],
[
"Ma",
"Ke",
""
],
[
"Liang",
"Shixiao",
""
],
[
"Li",
"Xiaopeng",
""
],
[
"Qu",
"Xiaobo",
""
]
] | TITLE: ULTra-AV: A Unified Longitudinal Trajectory Dataset for Automated
Vehicle
ABSTRACT: Automated Vehicles (AVs) promise significant advances in transportation.
Critical to these improvements is understanding AVs' longitudinal behavior,
relying heavily on real-world trajectory data. Existing open-source trajectory
datasets of AV, however, often fall short in refinement, reliability, and
completeness, hindering effective performance metrics analysis and model
development. This study addresses these challenges by creating a Unified
Longitudinal TRAjectory dataset for AVs (Ultra-AV) to analyze their microscopic
longitudinal driving behaviors. This dataset compiles data from 13 distinct
sources, encompassing various AV types, test sites, and experiment scenarios.
We established a three-step data processing: 1. extraction of longitudinal
trajectory data, 2. general data cleaning, and 3. data-specific cleaning to
obtain the longitudinal trajectory data and car-following trajectory data. The
validity of the processed data is affirmed through performance evaluations
across safety, mobility, stability, and sustainability, along with an analysis
of the relationships between variables in car-following models. Our work not
only furnishes researchers with standardized data and metrics for longitudinal
AV behavior studies but also sets guidelines for data collection and model
development.
|
2406.17911 | Kun Zhao | Kun Zhao, Chenghao Xiao, Sixing Yan, William K. Cheung, Kai Ye, Noura
Al Moubayed, Liang Zhan, Chenghua Lin | X-ray Made Simple: Lay Radiology Report Generation and Robust Evaluation | This paper has substantial data and conceptual changes since release
that go beyond simple updating the existing one. As a result, the authors
have changed and we need to re-coordinate and reach consensus. So we decide
to withdraw it | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radiology Report Generation (RRG) has advanced considerably with the
development of multimodal generative models. Despite the progress, the field
still faces significant challenges in evaluation, as existing metrics lack
robustness and fairness. We reveal that, RRG with high performance on existing
lexical-based metrics (e.g. BLEU) might be more of a mirage - a model can get a
high BLEU only by learning the template of reports. This has become a pressing
issue for RRG due to the highly patternized nature of these reports. In
addition, standard radiology reports are often highly technical. Helping
patients understand these reports is crucial from a patient's perspective, yet
this has been largely overlooked in previous work. In this work, we
un-intuitively approach these problems by proposing the Layman's RRG framework
that can systematically improve RRG with day-to-day language. Specifically, our
framework first contributes a translated Layman's terms dataset. Building upon
the dataset, we then propose a semantics-based evaluation method, which is
effective in mitigating the inflated numbers of BLEU and provides more robust
evaluation. We show that training on the layman's terms dataset encourages
models to focus on the semantics of the reports, as opposed to overfitting to
learning the report templates. Last, we reveal a promising scaling law between
the number of training examples and semantics gain provided by our dataset,
compared to the inverse pattern brought by the original formats.
| [
{
"version": "v1",
"created": "Tue, 25 Jun 2024 19:52:01 GMT"
},
{
"version": "v2",
"created": "Sun, 30 Jun 2024 21:53:56 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Oct 2024 21:57:58 GMT"
},
{
"version": "v4",
"created": "Fri, 21 Feb 2025 22:52:11 GMT"
},
{
"version": "v5",
"created": "Fri, 14 Mar 2025 14:44:32 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Zhao",
"Kun",
""
],
[
"Xiao",
"Chenghao",
""
],
[
"Yan",
"Sixing",
""
],
[
"Cheung",
"William K.",
""
],
[
"Ye",
"Kai",
""
],
[
"Moubayed",
"Noura Al",
""
],
[
"Zhan",
"Liang",
""
],
[
"Lin",
"Chenghua",
""
]
] | TITLE: X-ray Made Simple: Lay Radiology Report Generation and Robust Evaluation
ABSTRACT: Radiology Report Generation (RRG) has advanced considerably with the
development of multimodal generative models. Despite the progress, the field
still faces significant challenges in evaluation, as existing metrics lack
robustness and fairness. We reveal that, RRG with high performance on existing
lexical-based metrics (e.g. BLEU) might be more of a mirage - a model can get a
high BLEU only by learning the template of reports. This has become a pressing
issue for RRG due to the highly patternized nature of these reports. In
addition, standard radiology reports are often highly technical. Helping
patients understand these reports is crucial from a patient's perspective, yet
this has been largely overlooked in previous work. In this work, we
un-intuitively approach these problems by proposing the Layman's RRG framework
that can systematically improve RRG with day-to-day language. Specifically, our
framework first contributes a translated Layman's terms dataset. Building upon
the dataset, we then propose a semantics-based evaluation method, which is
effective in mitigating the inflated numbers of BLEU and provides more robust
evaluation. We show that training on the layman's terms dataset encourages
models to focus on the semantics of the reports, as opposed to overfitting to
learning the report templates. Last, we reveal a promising scaling law between
the number of training examples and semantics gain provided by our dataset,
compared to the inverse pattern brought by the original formats.
|
2407.11211 | Philipp Allgeuer | Philipp Allgeuer and Kyra Ahrens and Stefan Wermter | Unconstrained Open Vocabulary Image Classification: Zero-Shot Transfer
from Text to Image via CLIP Inversion | Published at WACV 2025 | Winter Conference on Applications of Computer Vision (WACV), 2025,
pp. 8206-8217 | null | null | cs.CV cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | We introduce NOVIC, an innovative real-time uNconstrained Open Vocabulary
Image Classifier that uses an autoregressive transformer to generatively output
classification labels as language. Leveraging the extensive knowledge of CLIP
models, NOVIC harnesses the embedding space to enable zero-shot transfer from
pure text to images. Traditional CLIP models, despite their ability for open
vocabulary classification, require an exhaustive prompt of potential class
labels, restricting their application to images of known content or context. To
address this, we propose an "object decoder" model that is trained on a
large-scale 92M-target dataset of templated object noun sets and LLM-generated
captions to always output the object noun in question. This effectively inverts
the CLIP text encoder and allows textual object labels from essentially the
entire English language to be generated directly from image-derived embedding
vectors, without requiring any a priori knowledge of the potential content of
an image, and without any label biases. The trained decoders are tested on a
mix of manually and web-curated datasets, as well as standard image
classification benchmarks, and achieve fine-grained prompt-free prediction
scores of up to 87.5%, a strong result considering the model must work for any
conceivable image and without any contextual clues.
| [
{
"version": "v1",
"created": "Mon, 15 Jul 2024 19:53:02 GMT"
},
{
"version": "v2",
"created": "Wed, 17 Jul 2024 22:23:42 GMT"
},
{
"version": "v3",
"created": "Mon, 18 Nov 2024 14:43:38 GMT"
},
{
"version": "v4",
"created": "Tue, 26 Nov 2024 09:28:35 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Allgeuer",
"Philipp",
""
],
[
"Ahrens",
"Kyra",
""
],
[
"Wermter",
"Stefan",
""
]
] | TITLE: Unconstrained Open Vocabulary Image Classification: Zero-Shot Transfer
from Text to Image via CLIP Inversion
ABSTRACT: We introduce NOVIC, an innovative real-time uNconstrained Open Vocabulary
Image Classifier that uses an autoregressive transformer to generatively output
classification labels as language. Leveraging the extensive knowledge of CLIP
models, NOVIC harnesses the embedding space to enable zero-shot transfer from
pure text to images. Traditional CLIP models, despite their ability for open
vocabulary classification, require an exhaustive prompt of potential class
labels, restricting their application to images of known content or context. To
address this, we propose an "object decoder" model that is trained on a
large-scale 92M-target dataset of templated object noun sets and LLM-generated
captions to always output the object noun in question. This effectively inverts
the CLIP text encoder and allows textual object labels from essentially the
entire English language to be generated directly from image-derived embedding
vectors, without requiring any a priori knowledge of the potential content of
an image, and without any label biases. The trained decoders are tested on a
mix of manually and web-curated datasets, as well as standard image
classification benchmarks, and achieve fine-grained prompt-free prediction
scores of up to 87.5%, a strong result considering the model must work for any
conceivable image and without any contextual clues.
|
2408.12928 | An-Lan Wang | An-Lan Wang, Bin Shan, Wei Shi, Kun-Yu Lin, Xiang Fei, Guozhi Tang,
Lei Liao, Can Huang, Jingqun Tang, Wei-Shi Zheng | ParGo: Bridging Vision-Language with Partial and Global Views | Accepted by AAAI 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | This work presents ParGo, a novel Partial-Global projector designed to
connect the vision and language modalities for Multimodal Large Language Models
(MLLMs). Unlike previous works that rely on global attention-based projectors,
our ParGo bridges the representation gap between the separately pre-trained
vision encoders and the LLMs by integrating global and partial views, which
alleviates the overemphasis on prominent regions. To facilitate the effective
training of ParGo, we collect a large-scale detail-captioned image-text dataset
named ParGoCap-1M-PT, consisting of 1 million images paired with high-quality
captions. Extensive experiments on several MLLM benchmarks demonstrate the
effectiveness of our ParGo, highlighting its superiority in aligning vision and
language modalities. Compared to conventional Q-Former projector, our ParGo
achieves an improvement of 259.96 in MME benchmark. Furthermore, our
experiments reveal that ParGo significantly outperforms other projectors,
particularly in tasks that emphasize detail perception ability.
| [
{
"version": "v1",
"created": "Fri, 23 Aug 2024 09:14:58 GMT"
},
{
"version": "v2",
"created": "Tue, 7 Jan 2025 09:39:15 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 08:48:51 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Wang",
"An-Lan",
""
],
[
"Shan",
"Bin",
""
],
[
"Shi",
"Wei",
""
],
[
"Lin",
"Kun-Yu",
""
],
[
"Fei",
"Xiang",
""
],
[
"Tang",
"Guozhi",
""
],
[
"Liao",
"Lei",
""
],
[
"Huang",
"Can",
""
],
[
"Tang",
"Jingqun",
""
],
[
"Zheng",
"Wei-Shi",
""
]
] | TITLE: ParGo: Bridging Vision-Language with Partial and Global Views
ABSTRACT: This work presents ParGo, a novel Partial-Global projector designed to
connect the vision and language modalities for Multimodal Large Language Models
(MLLMs). Unlike previous works that rely on global attention-based projectors,
our ParGo bridges the representation gap between the separately pre-trained
vision encoders and the LLMs by integrating global and partial views, which
alleviates the overemphasis on prominent regions. To facilitate the effective
training of ParGo, we collect a large-scale detail-captioned image-text dataset
named ParGoCap-1M-PT, consisting of 1 million images paired with high-quality
captions. Extensive experiments on several MLLM benchmarks demonstrate the
effectiveness of our ParGo, highlighting its superiority in aligning vision and
language modalities. Compared to conventional Q-Former projector, our ParGo
achieves an improvement of 259.96 in MME benchmark. Furthermore, our
experiments reveal that ParGo significantly outperforms other projectors,
particularly in tasks that emphasize detail perception ability.
|
2409.00138 | Yijia Shao | Yijia Shao, Tianshi Li, Weiyan Shi, Yanchen Liu, Diyi Yang | PrivacyLens: Evaluating Privacy Norm Awareness of Language Models in
Action | NeurIPS 2024 Datasets and Benchmarks Track | null | null | null | cs.CL cs.AI cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As language models (LMs) are widely utilized in personalized communication
scenarios (e.g., sending emails, writing social media posts) and endowed with a
certain level of agency, ensuring they act in accordance with the contextual
privacy norms becomes increasingly critical. However, quantifying the privacy
norm awareness of LMs and the emerging privacy risk in LM-mediated
communication is challenging due to (1) the contextual and long-tailed nature
of privacy-sensitive cases, and (2) the lack of evaluation approaches that
capture realistic application scenarios. To address these challenges, we
propose PrivacyLens, a novel framework designed to extend privacy-sensitive
seeds into expressive vignettes and further into agent trajectories, enabling
multi-level evaluation of privacy leakage in LM agents' actions. We instantiate
PrivacyLens with a collection of privacy norms grounded in privacy literature
and crowdsourced seeds. Using this dataset, we reveal a discrepancy between LM
performance in answering probing questions and their actual behavior when
executing user instructions in an agent setup. State-of-the-art LMs, like GPT-4
and Llama-3-70B, leak sensitive information in 25.68% and 38.69% of cases, even
when prompted with privacy-enhancing instructions. We also demonstrate the
dynamic nature of PrivacyLens by extending each seed into multiple trajectories
to red-team LM privacy leakage risk. Dataset and code are available at
https://github.com/SALT-NLP/PrivacyLens.
| [
{
"version": "v1",
"created": "Thu, 29 Aug 2024 17:58:38 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Oct 2024 04:43:40 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 06:03:20 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Shao",
"Yijia",
""
],
[
"Li",
"Tianshi",
""
],
[
"Shi",
"Weiyan",
""
],
[
"Liu",
"Yanchen",
""
],
[
"Yang",
"Diyi",
""
]
] | TITLE: PrivacyLens: Evaluating Privacy Norm Awareness of Language Models in
Action
ABSTRACT: As language models (LMs) are widely utilized in personalized communication
scenarios (e.g., sending emails, writing social media posts) and endowed with a
certain level of agency, ensuring they act in accordance with the contextual
privacy norms becomes increasingly critical. However, quantifying the privacy
norm awareness of LMs and the emerging privacy risk in LM-mediated
communication is challenging due to (1) the contextual and long-tailed nature
of privacy-sensitive cases, and (2) the lack of evaluation approaches that
capture realistic application scenarios. To address these challenges, we
propose PrivacyLens, a novel framework designed to extend privacy-sensitive
seeds into expressive vignettes and further into agent trajectories, enabling
multi-level evaluation of privacy leakage in LM agents' actions. We instantiate
PrivacyLens with a collection of privacy norms grounded in privacy literature
and crowdsourced seeds. Using this dataset, we reveal a discrepancy between LM
performance in answering probing questions and their actual behavior when
executing user instructions in an agent setup. State-of-the-art LMs, like GPT-4
and Llama-3-70B, leak sensitive information in 25.68% and 38.69% of cases, even
when prompted with privacy-enhancing instructions. We also demonstrate the
dynamic nature of PrivacyLens by extending each seed into multiple trajectories
to red-team LM privacy leakage risk. Dataset and code are available at
https://github.com/SALT-NLP/PrivacyLens.
|
2409.02465 | Zhe Xu | Zhe Xu, Jiasheng Ye, Xiaoran Liu, Xiangyang Liu, Tianxiang Sun,
Zhigeng Liu, Qipeng Guo, Linlin Li, Qun Liu, Xuanjing Huang, Xipeng Qiu | DetectiveQA: Evaluating Long-Context Reasoning on Detective Novels | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, significant efforts have been devoted to enhancing the long-context
capabilities of Large Language Models (LLMs), particularly in long-context
reasoning. To facilitate this research, we propose \textbf{DetectiveQA}, a
dataset specifically designed for narrative reasoning within long contexts. We
leverage detective novels, averaging over 100k tokens, to create a dataset
containing 1200 human-annotated questions in both Chinese and English, each
paired with corresponding reference reasoning steps. Furthermore, we introduce
a step-wise reasoning metric, which enhances the evaluation of LLMs' reasoning
processes. We validate our approach and evaluate the mainstream LLMs, including
GPT-4, Claude, and LLaMA, revealing persistent long-context reasoning
challenges and demonstrating their evidence-retrieval challenges. Our findings
offer valuable insights into the study of long-context reasoning and lay the
base for more rigorous evaluations.
| [
{
"version": "v1",
"created": "Wed, 4 Sep 2024 06:28:22 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 08:44:06 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Xu",
"Zhe",
""
],
[
"Ye",
"Jiasheng",
""
],
[
"Liu",
"Xiaoran",
""
],
[
"Liu",
"Xiangyang",
""
],
[
"Sun",
"Tianxiang",
""
],
[
"Liu",
"Zhigeng",
""
],
[
"Guo",
"Qipeng",
""
],
[
"Li",
"Linlin",
""
],
[
"Liu",
"Qun",
""
],
[
"Huang",
"Xuanjing",
""
],
[
"Qiu",
"Xipeng",
""
]
] | TITLE: DetectiveQA: Evaluating Long-Context Reasoning on Detective Novels
ABSTRACT: Recently, significant efforts have been devoted to enhancing the long-context
capabilities of Large Language Models (LLMs), particularly in long-context
reasoning. To facilitate this research, we propose \textbf{DetectiveQA}, a
dataset specifically designed for narrative reasoning within long contexts. We
leverage detective novels, averaging over 100k tokens, to create a dataset
containing 1200 human-annotated questions in both Chinese and English, each
paired with corresponding reference reasoning steps. Furthermore, we introduce
a step-wise reasoning metric, which enhances the evaluation of LLMs' reasoning
processes. We validate our approach and evaluate the mainstream LLMs, including
GPT-4, Claude, and LLaMA, revealing persistent long-context reasoning
challenges and demonstrating their evidence-retrieval challenges. Our findings
offer valuable insights into the study of long-context reasoning and lay the
base for more rigorous evaluations.
|
2409.03277 | Zhengzhuo Xu | Zhengzhuo Xu, Bowen Qu, Yiyan Qi, Sinan Du, Chengjin Xu, Chun Yuan,
Jian Guo | ChartMoE: Mixture of Diversely Aligned Expert Connector for Chart
Understanding | null | null | null | null | cs.AI cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | Automatic chart understanding is crucial for content comprehension and
document parsing. Multimodal Large Language Models (MLLMs) have demonstrated
remarkable capabilities in chart understanding through domain-specific
alignment and fine-tuning. However, current MLLMs still struggle to provide
faithful data and reliable analysis only based on charts. To address it, we
propose ChartMoE, which employs the Mixture of Expert (MoE) architecture to
replace the traditional linear projector to bridge the modality gap.
Specifically, we train several linear connectors through distinct alignment
tasks, which are utilized as the foundational initialization parameters for
different experts. Additionally, we introduce ChartMoE-Align, a dataset with
nearly 1 million chart-table-JSON-code quadruples to conduct three alignment
tasks (chart-table/JSON/code). Combined with the vanilla connector, we
initialize different experts diversely and adopt high-quality knowledge
learning to further refine the MoE connector and LLM parameters. Extensive
experiments demonstrate the effectiveness of the MoE connector and our
initialization strategy, e.g., ChartMoE improves the accuracy of the previous
state-of-the-art from 80.48\% to 84.64\% on the ChartQA benchmark.
| [
{
"version": "v1",
"created": "Thu, 5 Sep 2024 06:41:02 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Feb 2025 17:22:34 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 03:19:00 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Xu",
"Zhengzhuo",
""
],
[
"Qu",
"Bowen",
""
],
[
"Qi",
"Yiyan",
""
],
[
"Du",
"Sinan",
""
],
[
"Xu",
"Chengjin",
""
],
[
"Yuan",
"Chun",
""
],
[
"Guo",
"Jian",
""
]
] | TITLE: ChartMoE: Mixture of Diversely Aligned Expert Connector for Chart
Understanding
ABSTRACT: Automatic chart understanding is crucial for content comprehension and
document parsing. Multimodal Large Language Models (MLLMs) have demonstrated
remarkable capabilities in chart understanding through domain-specific
alignment and fine-tuning. However, current MLLMs still struggle to provide
faithful data and reliable analysis only based on charts. To address it, we
propose ChartMoE, which employs the Mixture of Expert (MoE) architecture to
replace the traditional linear projector to bridge the modality gap.
Specifically, we train several linear connectors through distinct alignment
tasks, which are utilized as the foundational initialization parameters for
different experts. Additionally, we introduce ChartMoE-Align, a dataset with
nearly 1 million chart-table-JSON-code quadruples to conduct three alignment
tasks (chart-table/JSON/code). Combined with the vanilla connector, we
initialize different experts diversely and adopt high-quality knowledge
learning to further refine the MoE connector and LLM parameters. Extensive
experiments demonstrate the effectiveness of the MoE connector and our
initialization strategy, e.g., ChartMoE improves the accuracy of the previous
state-of-the-art from 80.48\% to 84.64\% on the ChartQA benchmark.
|
2409.06316 | Daniel Rose | Daniel Rose, Oliver Wieder, Thomas Seidel, Thierry Langer | PharmacoMatch: Efficient 3D Pharmacophore Screening via Neural Subgraph
Matching | null | null | null | null | cs.LG cs.AI q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | The increasing size of screening libraries poses a significant challenge for
the development of virtual screening methods for drug discovery, necessitating
a re-evaluation of traditional approaches in the era of big data. Although 3D
pharmacophore screening remains a prevalent technique, its application to very
large datasets is limited by the computational cost associated with matching
query pharmacophores to database molecules. In this study, we introduce
PharmacoMatch, a novel contrastive learning approach based on neural subgraph
matching. Our method reinterprets pharmacophore screening as an approximate
subgraph matching problem and enables efficient querying of conformational
databases by encoding query-target relationships in the embedding space. We
conduct comprehensive investigations of the learned representations and
evaluate PharmacoMatch as pre-screening tool in a zero-shot setting. We
demonstrate significantly shorter runtimes and comparable performance metrics
to existing solutions, providing a promising speed-up for screening very large
datasets.
| [
{
"version": "v1",
"created": "Tue, 10 Sep 2024 08:17:06 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 09:51:43 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Rose",
"Daniel",
""
],
[
"Wieder",
"Oliver",
""
],
[
"Seidel",
"Thomas",
""
],
[
"Langer",
"Thierry",
""
]
] | TITLE: PharmacoMatch: Efficient 3D Pharmacophore Screening via Neural Subgraph
Matching
ABSTRACT: The increasing size of screening libraries poses a significant challenge for
the development of virtual screening methods for drug discovery, necessitating
a re-evaluation of traditional approaches in the era of big data. Although 3D
pharmacophore screening remains a prevalent technique, its application to very
large datasets is limited by the computational cost associated with matching
query pharmacophores to database molecules. In this study, we introduce
PharmacoMatch, a novel contrastive learning approach based on neural subgraph
matching. Our method reinterprets pharmacophore screening as an approximate
subgraph matching problem and enables efficient querying of conformational
databases by encoding query-target relationships in the embedding space. We
conduct comprehensive investigations of the learned representations and
evaluate PharmacoMatch as pre-screening tool in a zero-shot setting. We
demonstrate significantly shorter runtimes and comparable performance metrics
to existing solutions, providing a promising speed-up for screening very large
datasets.
|
2409.15397 | Danijel Korzinek | Nikola Ljube\v{s}i\'c, Peter Rupnik and Danijel Kor\v{z}inek | The ParlaSpeech Collection of Automatically Generated Speech and Text
Datasets from Parliamentary Proceedings | Submitted to SPECOM 2024 | null | 10.1007/978-3-031-77961-9_10 | null | eess.AS cs.CL cs.LG cs.SD | http://creativecommons.org/licenses/by/4.0/ | Recent significant improvements in speech and language technologies come both
from self-supervised approaches over raw language data as well as various types
of explicit supervision. To ensure high-quality processing of spoken data, the
most useful type of explicit supervision is still the alignment between the
speech signal and its corresponding text transcript, which is a data type that
is not available for many languages. In this paper, we present our approach to
building large and open speech-and-text-aligned datasets of less-resourced
languages based on transcripts of parliamentary proceedings and their
recordings. Our starting point are the ParlaMint comparable corpora of
transcripts of parliamentary proceedings of 26 national European parliaments.
In the pilot run on expanding the ParlaMint corpora with aligned publicly
available recordings, we focus on three Slavic languages, namely Croatian,
Polish, and Serbian. The main challenge of our approach is the lack of any
global alignment between the ParlaMint texts and the available recordings, as
well as the sometimes varying data order in each of the modalities, which
requires a novel approach in aligning long sequences of text and audio in a
large search space. The results of this pilot run are three high-quality
datasets that span more than 5,000 hours of speech and accompanying text
transcripts. Although these datasets already make a huge difference in the
availability of spoken and textual data for the three languages, we want to
emphasize the potential of the presented approach in building similar datasets
for many more languages.
| [
{
"version": "v1",
"created": "Mon, 23 Sep 2024 10:12:18 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Nov 2024 12:50:50 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Ljubešić",
"Nikola",
""
],
[
"Rupnik",
"Peter",
""
],
[
"Koržinek",
"Danijel",
""
]
] | TITLE: The ParlaSpeech Collection of Automatically Generated Speech and Text
Datasets from Parliamentary Proceedings
ABSTRACT: Recent significant improvements in speech and language technologies come both
from self-supervised approaches over raw language data as well as various types
of explicit supervision. To ensure high-quality processing of spoken data, the
most useful type of explicit supervision is still the alignment between the
speech signal and its corresponding text transcript, which is a data type that
is not available for many languages. In this paper, we present our approach to
building large and open speech-and-text-aligned datasets of less-resourced
languages based on transcripts of parliamentary proceedings and their
recordings. Our starting point are the ParlaMint comparable corpora of
transcripts of parliamentary proceedings of 26 national European parliaments.
In the pilot run on expanding the ParlaMint corpora with aligned publicly
available recordings, we focus on three Slavic languages, namely Croatian,
Polish, and Serbian. The main challenge of our approach is the lack of any
global alignment between the ParlaMint texts and the available recordings, as
well as the sometimes varying data order in each of the modalities, which
requires a novel approach in aligning long sequences of text and audio in a
large search space. The results of this pilot run are three high-quality
datasets that span more than 5,000 hours of speech and accompanying text
transcripts. Although these datasets already make a huge difference in the
availability of spoken and textual data for the three languages, we want to
emphasize the potential of the presented approach in building similar datasets
for many more languages.
|
2409.15784 | Sung Yun Lee | Sung Yun Lee, Do Hyung Cho, Chulho Jung, Daeho Sung, Daewoong Nam,
Sangsoo Kim, Changyong Song | Deep-learning real-time phase retrieval of imperfect diffraction
patterns from X-ray free-electron lasers | null | null | 10.1038/s41524-025-01569-7 | null | physics.app-ph cond-mat.mtrl-sci cs.LG physics.optics | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Machine learning is attracting surging interest across nearly all scientific
areas by enabling the analysis of large datasets and the extraction of
scientific information from incomplete data. Data-driven science is rapidly
growing, especially in X-ray methodologies, where advanced light sources and
detection technologies accumulate vast amounts of data that exceed meticulous
human inspection capabilities. Despite the increasing demands, the full
application of machine learning has been hindered by the need for data-specific
optimizations. In this study, we introduce a new deep-learning-based phase
retrieval method for imperfect diffraction data. This method provides robust
phase retrieval for simulated data and performs well on weak-signal
single-pulse diffraction data from X-ray free-electron lasers. Moreover, the
method significantly reduces data processing time, facilitating real-time image
reconstructions that are crucial for high-repetition-rate data acquisition.
Thus, this approach offers a reliable solution to the phase problem and is
expected to be widely adopted across various research areas.
| [
{
"version": "v1",
"created": "Tue, 24 Sep 2024 06:28:25 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Lee",
"Sung Yun",
""
],
[
"Cho",
"Do Hyung",
""
],
[
"Jung",
"Chulho",
""
],
[
"Sung",
"Daeho",
""
],
[
"Nam",
"Daewoong",
""
],
[
"Kim",
"Sangsoo",
""
],
[
"Song",
"Changyong",
""
]
] | TITLE: Deep-learning real-time phase retrieval of imperfect diffraction
patterns from X-ray free-electron lasers
ABSTRACT: Machine learning is attracting surging interest across nearly all scientific
areas by enabling the analysis of large datasets and the extraction of
scientific information from incomplete data. Data-driven science is rapidly
growing, especially in X-ray methodologies, where advanced light sources and
detection technologies accumulate vast amounts of data that exceed meticulous
human inspection capabilities. Despite the increasing demands, the full
application of machine learning has been hindered by the need for data-specific
optimizations. In this study, we introduce a new deep-learning-based phase
retrieval method for imperfect diffraction data. This method provides robust
phase retrieval for simulated data and performs well on weak-signal
single-pulse diffraction data from X-ray free-electron lasers. Moreover, the
method significantly reduces data processing time, facilitating real-time image
reconstructions that are crucial for high-repetition-rate data acquisition.
Thus, this approach offers a reliable solution to the phase problem and is
expected to be widely adopted across various research areas.
|
2410.02086 | Minoh Jeong | Minoh Jeong, Min Namgung, Zae Myung Kim, Dongyeop Kang, Yao-Yi Chiang,
Alfred Hero | Anchors Aweigh! Sail for Optimal Unified Multi-Modal Representations | null | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A unified representation space in multi-modal learning is essential for
effectively integrating diverse data sources, such as text, images, and audio,
to enhance efficiency and performance across various downstream tasks. Recent
binding methods, such as ImageBind (Girdhar et al., 2023), typically rely on a
single, fixed anchor modality for aligning multi-modal data. We mathematically
analyze these fixed anchor binding method and uncover significant limitations:
(1) over-reliance on the choice of the anchor modality, (2) inadequate capture
of intra-modal information, and (3) failure to account for cross-modal
correlation among non-anchored modalities. To address these issues, we propose
the need for adaptive anchor binding methods, exemplified by our framework
CentroBind. The proposed method uses adaptively adjustable centroid-based
anchors generated from all available modalities, leading to a balanced and rich
representation space. We theoretically demonstrate that our approach captures
three critical properties of multi-modal learning -- intra-modal learning,
inter-modal learning, and multi-modal alignment -- while constructing a unified
representation that spans all modalities. Experiments on both synthetic and
real-world datasets show that adaptive anchor methods such as CentroBind
consistently outperform fixed anchor binding methods, verifying our analysis.
| [
{
"version": "v1",
"created": "Wed, 2 Oct 2024 23:19:23 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 16:36:53 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Jeong",
"Minoh",
""
],
[
"Namgung",
"Min",
""
],
[
"Kim",
"Zae Myung",
""
],
[
"Kang",
"Dongyeop",
""
],
[
"Chiang",
"Yao-Yi",
""
],
[
"Hero",
"Alfred",
""
]
] | TITLE: Anchors Aweigh! Sail for Optimal Unified Multi-Modal Representations
ABSTRACT: A unified representation space in multi-modal learning is essential for
effectively integrating diverse data sources, such as text, images, and audio,
to enhance efficiency and performance across various downstream tasks. Recent
binding methods, such as ImageBind (Girdhar et al., 2023), typically rely on a
single, fixed anchor modality for aligning multi-modal data. We mathematically
analyze these fixed anchor binding method and uncover significant limitations:
(1) over-reliance on the choice of the anchor modality, (2) inadequate capture
of intra-modal information, and (3) failure to account for cross-modal
correlation among non-anchored modalities. To address these issues, we propose
the need for adaptive anchor binding methods, exemplified by our framework
CentroBind. The proposed method uses adaptively adjustable centroid-based
anchors generated from all available modalities, leading to a balanced and rich
representation space. We theoretically demonstrate that our approach captures
three critical properties of multi-modal learning -- intra-modal learning,
inter-modal learning, and multi-modal alignment -- while constructing a unified
representation that spans all modalities. Experiments on both synthetic and
real-world datasets show that adaptive anchor methods such as CentroBind
consistently outperform fixed anchor binding methods, verifying our analysis.
|
2410.02603 | Fantine Huot | Fantine Huot, Reinald Kim Amplayo, Jennimaria Palomaki, Alice Shoshana
Jakobovits, Elizabeth Clark, Mirella Lapata | Agents' Room: Narrative Generation through Multi-step Collaboration | Published as a conference paper at ICLR 2025 | null | null | null | cs.CL cs.LG cs.MA | http://creativecommons.org/licenses/by/4.0/ | Writing compelling fiction is a multifaceted process combining elements such
as crafting a plot, developing interesting characters, and using evocative
language. While large language models (LLMs) show promise for story writing,
they currently rely heavily on intricate prompting, which limits their use. We
propose Agents' Room, a generation framework inspired by narrative theory, that
decomposes narrative writing into subtasks tackled by specialized agents. To
illustrate our method, we introduce Tell Me A Story, a high-quality dataset of
complex writing prompts and human-written stories, and a novel evaluation
framework designed specifically for assessing long narratives. We show that
Agents' Room generates stories that are preferred by expert evaluators over
those produced by baseline systems by leveraging collaboration and
specialization to decompose the complex story writing task into tractable
components. We provide extensive analysis with automated and human-based
metrics of the generated output.
| [
{
"version": "v1",
"created": "Thu, 3 Oct 2024 15:44:42 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 17:09:03 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Huot",
"Fantine",
""
],
[
"Amplayo",
"Reinald Kim",
""
],
[
"Palomaki",
"Jennimaria",
""
],
[
"Jakobovits",
"Alice Shoshana",
""
],
[
"Clark",
"Elizabeth",
""
],
[
"Lapata",
"Mirella",
""
]
] | TITLE: Agents' Room: Narrative Generation through Multi-step Collaboration
ABSTRACT: Writing compelling fiction is a multifaceted process combining elements such
as crafting a plot, developing interesting characters, and using evocative
language. While large language models (LLMs) show promise for story writing,
they currently rely heavily on intricate prompting, which limits their use. We
propose Agents' Room, a generation framework inspired by narrative theory, that
decomposes narrative writing into subtasks tackled by specialized agents. To
illustrate our method, we introduce Tell Me A Story, a high-quality dataset of
complex writing prompts and human-written stories, and a novel evaluation
framework designed specifically for assessing long narratives. We show that
Agents' Room generates stories that are preferred by expert evaluators over
those produced by baseline systems by leveraging collaboration and
specialization to decompose the complex story writing task into tractable
components. We provide extensive analysis with automated and human-based
metrics of the generated output.
|
2410.03461 | Tobias Leemann | Tobias Leemann, Periklis Petridis, Giuseppe Vietri, Dionysis
Manousakas, Aaron Roth, Sergul Aydore | Auto-GDA: Automatic Domain Adaptation for Efficient Grounding
Verification in Retrieval-Augmented Generation | null | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While retrieval-augmented generation (RAG) has been shown to enhance
factuality of large language model (LLM) outputs, LLMs still suffer from
hallucination, generating incorrect or irrelevant information. A common
detection strategy involves prompting the LLM again to assess whether its
response is grounded in the retrieved evidence, but this approach is costly.
Alternatively, lightweight natural language inference (NLI) models for
efficient grounding verification can be used at inference time. While existing
pre-trained NLI models offer potential solutions, their performance remains
subpar compared to larger models on realistic RAG inputs. RAG inputs are more
complex than most datasets used for training NLI models and have
characteristics specific to the underlying knowledge base, requiring adaptation
of the NLI models to a specific target domain. Additionally, the lack of
labeled instances in the target domain makes supervised domain adaptation,
e.g., through fine-tuning, infeasible. To address these challenges, we
introduce Automatic Generative Domain Adaptation (Auto-GDA). Our framework
enables unsupervised domain adaptation through synthetic data generation.
Unlike previous methods that rely on handcrafted filtering and augmentation
strategies, Auto-GDA employs an iterative process to continuously improve the
quality of generated samples using weak labels from less efficient teacher
models and discrete optimization to select the most promising augmented
samples. Experimental results demonstrate the effectiveness of our approach,
with models fine-tuned on synthetic data using Auto-GDA often surpassing the
performance of the teacher model and reaching the performance level of LLMs at
10% of their computational cost.
| [
{
"version": "v1",
"created": "Fri, 4 Oct 2024 14:21:27 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 17:27:00 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Leemann",
"Tobias",
""
],
[
"Petridis",
"Periklis",
""
],
[
"Vietri",
"Giuseppe",
""
],
[
"Manousakas",
"Dionysis",
""
],
[
"Roth",
"Aaron",
""
],
[
"Aydore",
"Sergul",
""
]
] | TITLE: Auto-GDA: Automatic Domain Adaptation for Efficient Grounding
Verification in Retrieval-Augmented Generation
ABSTRACT: While retrieval-augmented generation (RAG) has been shown to enhance
factuality of large language model (LLM) outputs, LLMs still suffer from
hallucination, generating incorrect or irrelevant information. A common
detection strategy involves prompting the LLM again to assess whether its
response is grounded in the retrieved evidence, but this approach is costly.
Alternatively, lightweight natural language inference (NLI) models for
efficient grounding verification can be used at inference time. While existing
pre-trained NLI models offer potential solutions, their performance remains
subpar compared to larger models on realistic RAG inputs. RAG inputs are more
complex than most datasets used for training NLI models and have
characteristics specific to the underlying knowledge base, requiring adaptation
of the NLI models to a specific target domain. Additionally, the lack of
labeled instances in the target domain makes supervised domain adaptation,
e.g., through fine-tuning, infeasible. To address these challenges, we
introduce Automatic Generative Domain Adaptation (Auto-GDA). Our framework
enables unsupervised domain adaptation through synthetic data generation.
Unlike previous methods that rely on handcrafted filtering and augmentation
strategies, Auto-GDA employs an iterative process to continuously improve the
quality of generated samples using weak labels from less efficient teacher
models and discrete optimization to select the most promising augmented
samples. Experimental results demonstrate the effectiveness of our approach,
with models fine-tuned on synthetic data using Auto-GDA often surpassing the
performance of the teacher model and reaching the performance level of LLMs at
10% of their computational cost.
|
2410.05111 | Qifeng Chen | Qifeng Chen, Sheng Yang, Sicong Du, Tao Tang, Peng Chen, Yuchi Huo | LiDAR-GS:Real-time LiDAR Re-Simulation using Gaussian Splatting | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present LiDAR-GS, a Gaussian Splatting (GS) method for real-time,
high-fidelity re-simulation of LiDAR scans in public urban road scenes. Recent
GS methods proposed for cameras have achieved significant advancements in
real-time rendering beyond Neural Radiance Fields (NeRF). However, applying GS
representation to LiDAR, an active 3D sensor type, poses several challenges
that must be addressed to preserve high accuracy and unique characteristics.
Specifically, LiDAR-GS designs a differentiable laser beam splatting, using
range-view representation for precise surface splatting by projecting lasers
onto micro cross-sections, effectively eliminating artifacts associated with
local affine approximations. Furthermore, LiDAR-GS leverages Neural Gaussian
Representation, which further integrate view-dependent clues, to represent key
LiDAR properties that are influenced by the incident direction and external
factors. Combining these practices with some essential adaptations, e.g.,
dynamic instances decomposition, LiDAR-GS succeeds in simultaneously
re-simulating depth, intensity, and ray-drop channels, achieving
state-of-the-art results in both rendering frame rate and quality on publically
available large scene datasets when compared with the methods using explicit
mesh or implicit NeRF. Our source code is publicly available at
https://www.github.com/cqf7419/LiDAR-GS.
| [
{
"version": "v1",
"created": "Mon, 7 Oct 2024 15:07:56 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 09:52:11 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Chen",
"Qifeng",
""
],
[
"Yang",
"Sheng",
""
],
[
"Du",
"Sicong",
""
],
[
"Tang",
"Tao",
""
],
[
"Chen",
"Peng",
""
],
[
"Huo",
"Yuchi",
""
]
] | TITLE: LiDAR-GS:Real-time LiDAR Re-Simulation using Gaussian Splatting
ABSTRACT: We present LiDAR-GS, a Gaussian Splatting (GS) method for real-time,
high-fidelity re-simulation of LiDAR scans in public urban road scenes. Recent
GS methods proposed for cameras have achieved significant advancements in
real-time rendering beyond Neural Radiance Fields (NeRF). However, applying GS
representation to LiDAR, an active 3D sensor type, poses several challenges
that must be addressed to preserve high accuracy and unique characteristics.
Specifically, LiDAR-GS designs a differentiable laser beam splatting, using
range-view representation for precise surface splatting by projecting lasers
onto micro cross-sections, effectively eliminating artifacts associated with
local affine approximations. Furthermore, LiDAR-GS leverages Neural Gaussian
Representation, which further integrate view-dependent clues, to represent key
LiDAR properties that are influenced by the incident direction and external
factors. Combining these practices with some essential adaptations, e.g.,
dynamic instances decomposition, LiDAR-GS succeeds in simultaneously
re-simulating depth, intensity, and ray-drop channels, achieving
state-of-the-art results in both rendering frame rate and quality on publically
available large scene datasets when compared with the methods using explicit
mesh or implicit NeRF. Our source code is publicly available at
https://www.github.com/cqf7419/LiDAR-GS.
|
2410.12953 | Aayush Agrawal | Aayush Agrawal, Aniruddh Sikdar, Rajini Makam, Suresh Sundaram, Suresh
Kumar Besai and Mahesh Gopi | Syn2Real Domain Generalization for Underwater Mine-like Object Detection
Using Side-Scan Sonar | 7 pages, 4 figures and 3 tables | null | 10.1109/LGRS.2025.3550037 | null | cs.LG cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Underwater mine detection with deep learning suffers from limitations due to
the scarcity of real-world data.
This scarcity leads to overfitting, where models perform well on training
data but poorly on unseen data. This paper proposes a Syn2Real (Synthetic to
Real) domain generalization approach using diffusion models to address this
challenge. We demonstrate that synthetic data generated with noise by DDPM and
DDIM models, even if not perfectly realistic, can effectively augment
real-world samples for training. The residual noise in the final sampled images
improves the model's ability to generalize to real-world data with inherent
noise and high variation. The baseline Mask-RCNN model when trained on a
combination of synthetic and original training datasets, exhibited
approximately a 60% increase in Average Precision (AP) compared to being
trained solely on the original training data. This significant improvement
highlights the potential of Syn2Real domain generalization for underwater mine
detection tasks.
| [
{
"version": "v1",
"created": "Wed, 16 Oct 2024 18:42:08 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Agrawal",
"Aayush",
""
],
[
"Sikdar",
"Aniruddh",
""
],
[
"Makam",
"Rajini",
""
],
[
"Sundaram",
"Suresh",
""
],
[
"Besai",
"Suresh Kumar",
""
],
[
"Gopi",
"Mahesh",
""
]
] | TITLE: Syn2Real Domain Generalization for Underwater Mine-like Object Detection
Using Side-Scan Sonar
ABSTRACT: Underwater mine detection with deep learning suffers from limitations due to
the scarcity of real-world data.
This scarcity leads to overfitting, where models perform well on training
data but poorly on unseen data. This paper proposes a Syn2Real (Synthetic to
Real) domain generalization approach using diffusion models to address this
challenge. We demonstrate that synthetic data generated with noise by DDPM and
DDIM models, even if not perfectly realistic, can effectively augment
real-world samples for training. The residual noise in the final sampled images
improves the model's ability to generalize to real-world data with inherent
noise and high variation. The baseline Mask-RCNN model when trained on a
combination of synthetic and original training datasets, exhibited
approximately a 60% increase in Average Precision (AP) compared to being
trained solely on the original training data. This significant improvement
highlights the potential of Syn2Real domain generalization for underwater mine
detection tasks.
|
2410.19150 | Abraham Israeli | Abraham Israeli, David Jurgens, Daniel Romero | A Test of Time: Predicting the Sustainable Success of Online
Collaboration in Wikipedia | null | null | null | null | cs.CY cs.CL cs.SI | http://creativecommons.org/licenses/by/4.0/ | The Internet has significantly expanded the potential for global
collaboration, allowing millions of users to contribute to collective projects
like Wikipedia. While prior work has assessed the success of online
collaborations, most approaches are time-agnostic, evaluating success without
considering its longevity. Research on the factors that ensure the long-term
preservation of high-quality standards in online collaboration is scarce. In
this study, we address this gap. We propose a novel metric, `Sustainable
Success,' which measures the ability of collaborative efforts to maintain their
quality over time. Using Wikipedia as a case study, we introduce the
SustainPedia dataset, which compiles data from over 40K Wikipedia articles,
including each article's sustainable success label and more than 300
explanatory features such as edit history, user experience, and team
composition. Using this dataset, we develop machine learning models to predict
the sustainable success of Wikipedia articles. Our best-performing model
achieves a high AU-ROC score of 0.88 on average. Our analysis reveals important
insights. For example, we find that the longer an article takes to be
recognized as high-quality, the more likely it is to maintain that status over
time (i.e., be sustainable). Additionally, user experience emerged as the most
critical predictor of sustainability. Our analysis provides insights into
broader collective actions beyond Wikipedia (e.g., online activism,
crowdsourced open-source software), where the same social dynamics that drive
success on Wikipedia might play a role. We make all data and code used for this
study publicly available for further research.
| [
{
"version": "v1",
"created": "Thu, 24 Oct 2024 20:42:53 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 17:47:49 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Israeli",
"Abraham",
""
],
[
"Jurgens",
"David",
""
],
[
"Romero",
"Daniel",
""
]
] | TITLE: A Test of Time: Predicting the Sustainable Success of Online
Collaboration in Wikipedia
ABSTRACT: The Internet has significantly expanded the potential for global
collaboration, allowing millions of users to contribute to collective projects
like Wikipedia. While prior work has assessed the success of online
collaborations, most approaches are time-agnostic, evaluating success without
considering its longevity. Research on the factors that ensure the long-term
preservation of high-quality standards in online collaboration is scarce. In
this study, we address this gap. We propose a novel metric, `Sustainable
Success,' which measures the ability of collaborative efforts to maintain their
quality over time. Using Wikipedia as a case study, we introduce the
SustainPedia dataset, which compiles data from over 40K Wikipedia articles,
including each article's sustainable success label and more than 300
explanatory features such as edit history, user experience, and team
composition. Using this dataset, we develop machine learning models to predict
the sustainable success of Wikipedia articles. Our best-performing model
achieves a high AU-ROC score of 0.88 on average. Our analysis reveals important
insights. For example, we find that the longer an article takes to be
recognized as high-quality, the more likely it is to maintain that status over
time (i.e., be sustainable). Additionally, user experience emerged as the most
critical predictor of sustainability. Our analysis provides insights into
broader collective actions beyond Wikipedia (e.g., online activism,
crowdsourced open-source software), where the same social dynamics that drive
success on Wikipedia might play a role. We make all data and code used for this
study publicly available for further research.
|
2410.21705 | Yuxun Qu | Yuxun Qu, Yongqiang Tang, Chenyang Zhang, Wensheng Zhang | AdaptGCD: Multi-Expert Adapter Tuning for Generalized Category Discovery | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Different from the traditional semi-supervised learning paradigm that is
constrained by the close-world assumption, Generalized Category Discovery (GCD)
presumes that the unlabeled dataset contains new categories not appearing in
the labeled set, and aims to not only classify old categories but also discover
new categories in the unlabeled data. Existing studies on GCD typically devote
to transferring the general knowledge from the self-supervised pretrained model
to the target GCD task via some fine-tuning strategies, such as partial tuning
and prompt learning. Nevertheless, these fine-tuning methods fail to make a
sound balance between the generalization capacity of pretrained backbone and
the adaptability to the GCD task. To fill this gap, in this paper, we propose a
novel adapter-tuning-based method named AdaptGCD, which is the first work to
introduce the adapter tuning into the GCD task and provides some key insights
expected to enlighten future research. Furthermore, considering the discrepancy
of supervision information between the old and new classes, a multi-expert
adapter structure equipped with a route assignment constraint is elaborately
devised, such that the data from old and new classes are separated into
different expert groups. Extensive experiments are conducted on 7 widely-used
datasets. The remarkable improvements in performance highlight the
effectiveness of our proposals.
| [
{
"version": "v1",
"created": "Tue, 29 Oct 2024 03:41:47 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 15:55:43 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Qu",
"Yuxun",
""
],
[
"Tang",
"Yongqiang",
""
],
[
"Zhang",
"Chenyang",
""
],
[
"Zhang",
"Wensheng",
""
]
] | TITLE: AdaptGCD: Multi-Expert Adapter Tuning for Generalized Category Discovery
ABSTRACT: Different from the traditional semi-supervised learning paradigm that is
constrained by the close-world assumption, Generalized Category Discovery (GCD)
presumes that the unlabeled dataset contains new categories not appearing in
the labeled set, and aims to not only classify old categories but also discover
new categories in the unlabeled data. Existing studies on GCD typically devote
to transferring the general knowledge from the self-supervised pretrained model
to the target GCD task via some fine-tuning strategies, such as partial tuning
and prompt learning. Nevertheless, these fine-tuning methods fail to make a
sound balance between the generalization capacity of pretrained backbone and
the adaptability to the GCD task. To fill this gap, in this paper, we propose a
novel adapter-tuning-based method named AdaptGCD, which is the first work to
introduce the adapter tuning into the GCD task and provides some key insights
expected to enlighten future research. Furthermore, considering the discrepancy
of supervision information between the old and new classes, a multi-expert
adapter structure equipped with a route assignment constraint is elaborately
devised, such that the data from old and new classes are separated into
different expert groups. Extensive experiments are conducted on 7 widely-used
datasets. The remarkable improvements in performance highlight the
effectiveness of our proposals.
|
2411.05609 | Cristiano Patr\'icio | Cristiano Patr\'icio, Lu\'is F. Teixeira, Jo\~ao C. Neves | A Two-Step Concept-Based Approach for Enhanced Interpretability and
Trust in Skin Lesion Diagnosis | Published in the Computational and Structural Biotechnology Journal | null | 10.1016/j.csbj.2025.02.013 | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | The main challenges hindering the adoption of deep learning-based systems in
clinical settings are the scarcity of annotated data and the lack of
interpretability and trust in these systems. Concept Bottleneck Models (CBMs)
offer inherent interpretability by constraining the final disease prediction on
a set of human-understandable concepts. However, this inherent interpretability
comes at the cost of greater annotation burden. Additionally, adding new
concepts requires retraining the entire system. In this work, we introduce a
novel two-step methodology that addresses both of these challenges. By
simulating the two stages of a CBM, we utilize a pretrained Vision Language
Model (VLM) to automatically predict clinical concepts, and an off-the-shelf
Large Language Model (LLM) to generate disease diagnoses based on the predicted
concepts. Furthermore, our approach supports test-time human intervention,
enabling corrections to predicted concepts, which improves final diagnoses and
enhances transparency in decision-making. We validate our approach on three
skin lesion datasets, demonstrating that it outperforms traditional CBMs and
state-of-the-art explainable methods, all without requiring any training and
utilizing only a few annotated examples. The code is available at
https://github.com/CristianoPatricio/2-step-concept-based-skin-diagnosis.
| [
{
"version": "v1",
"created": "Fri, 8 Nov 2024 14:52:42 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 09:51:44 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Patrício",
"Cristiano",
""
],
[
"Teixeira",
"Luís F.",
""
],
[
"Neves",
"João C.",
""
]
] | TITLE: A Two-Step Concept-Based Approach for Enhanced Interpretability and
Trust in Skin Lesion Diagnosis
ABSTRACT: The main challenges hindering the adoption of deep learning-based systems in
clinical settings are the scarcity of annotated data and the lack of
interpretability and trust in these systems. Concept Bottleneck Models (CBMs)
offer inherent interpretability by constraining the final disease prediction on
a set of human-understandable concepts. However, this inherent interpretability
comes at the cost of greater annotation burden. Additionally, adding new
concepts requires retraining the entire system. In this work, we introduce a
novel two-step methodology that addresses both of these challenges. By
simulating the two stages of a CBM, we utilize a pretrained Vision Language
Model (VLM) to automatically predict clinical concepts, and an off-the-shelf
Large Language Model (LLM) to generate disease diagnoses based on the predicted
concepts. Furthermore, our approach supports test-time human intervention,
enabling corrections to predicted concepts, which improves final diagnoses and
enhances transparency in decision-making. We validate our approach on three
skin lesion datasets, demonstrating that it outperforms traditional CBMs and
state-of-the-art explainable methods, all without requiring any training and
utilizing only a few annotated examples. The code is available at
https://github.com/CristianoPatricio/2-step-concept-based-skin-diagnosis.
|
2411.12523 | Rania Briq | Rania Briq, Jiangtao Wang, Stefan Kesselheim | Data Pruning in Generative Diffusion Models | null | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data pruning is the problem of identifying a core subset that is most
beneficial to training and discarding the remainder. While pruning strategies
are well studied for discriminative models like those used in classification,
little research has gone into their application to generative models.
Generative models aim to estimate the underlying distribution of the data, so
presumably they should benefit from larger datasets. In this work we aim to
shed light on the accuracy of this statement, specifically answer the question
of whether data pruning for generative diffusion models could have a positive
impact. Contrary to intuition, we show that eliminating redundant or noisy data
in large datasets is beneficial particularly when done strategically. We
experiment with several pruning methods including recent-state-of-art methods,
and evaluate over CelebA-HQ and ImageNet datasets. We demonstrate that a simple
clustering method outperforms other sophisticated and computationally demanding
methods. We further exhibit how we can leverage clustering to balance skewed
datasets in an unsupervised manner to allow fair sampling for underrepresented
populations in the data distribution, which is a crucial problem in generative
models.
| [
{
"version": "v1",
"created": "Tue, 19 Nov 2024 14:13:25 GMT"
},
{
"version": "v2",
"created": "Sat, 25 Jan 2025 12:41:48 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 13:11:28 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Briq",
"Rania",
""
],
[
"Wang",
"Jiangtao",
""
],
[
"Kesselheim",
"Stefan",
""
]
] | TITLE: Data Pruning in Generative Diffusion Models
ABSTRACT: Data pruning is the problem of identifying a core subset that is most
beneficial to training and discarding the remainder. While pruning strategies
are well studied for discriminative models like those used in classification,
little research has gone into their application to generative models.
Generative models aim to estimate the underlying distribution of the data, so
presumably they should benefit from larger datasets. In this work we aim to
shed light on the accuracy of this statement, specifically answer the question
of whether data pruning for generative diffusion models could have a positive
impact. Contrary to intuition, we show that eliminating redundant or noisy data
in large datasets is beneficial particularly when done strategically. We
experiment with several pruning methods including recent-state-of-art methods,
and evaluate over CelebA-HQ and ImageNet datasets. We demonstrate that a simple
clustering method outperforms other sophisticated and computationally demanding
methods. We further exhibit how we can leverage clustering to balance skewed
datasets in an unsupervised manner to allow fair sampling for underrepresented
populations in the data distribution, which is a crucial problem in generative
models.
|
2411.12590 | Neale Ratzlaff | Neale Ratzlaff, Matthew Lyle Olson, Musashi Hinck, Estelle Aflalo,
Shao-Yen Tseng, Vasudev Lal, Phillip Howard | Debias your Large Multi-Modal Model at Test-Time via Non-Contrastive
Visual Attribute Steering | 10 pages, 6 Figures, 8 Tables. arXiv admin note: text overlap with
arXiv:2410.13976 | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Multi-Modal Models (LMMs) have demonstrated impressive capabilities as
general-purpose chatbots able to engage in conversations about visual inputs.
However, their responses are influenced by societal biases present in their
training datasets, leading to undesirable differences in how the model responds
when presented with images depicting people of different demographics. In this
work, we propose a training-free debiasing framework for LMMs that intervenes
on the model's representations during text generation by constructing a
steering vector that reduces reference on protected attributes. Our framework
introduces two complementary methods: (1) a dataset-based approach that
constructs a steering vector by contrasting model activations on biased and
neutral inputs, and (2) a novel optimization-based approach designed for
low-resource settings, which constructs the steering vector using a single step
of gradient-based perturbation without requiring additional data. Our
experiments show that these interventions effectively reduce the propensity of
LMMs to generate text related to protected attributes while maintaining
sentiment and fluency. Furthermore, we demonstrate that debiased LMMs achieve
comparable accuracy to their unmodified counterparts on downstream tasks,
indicating that bias mitigation can be achieved without sacrificing model
performance.
| [
{
"version": "v1",
"created": "Fri, 15 Nov 2024 20:06:09 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Mar 2025 18:02:59 GMT"
}
] | 2025-03-17T00:00:00 | [
[
"Ratzlaff",
"Neale",
""
],
[
"Olson",
"Matthew Lyle",
""
],
[
"Hinck",
"Musashi",
""
],
[
"Aflalo",
"Estelle",
""
],
[
"Tseng",
"Shao-Yen",
""
],
[
"Lal",
"Vasudev",
""
],
[
"Howard",
"Phillip",
""
]
] | TITLE: Debias your Large Multi-Modal Model at Test-Time via Non-Contrastive
Visual Attribute Steering
ABSTRACT: Large Multi-Modal Models (LMMs) have demonstrated impressive capabilities as
general-purpose chatbots able to engage in conversations about visual inputs.
However, their responses are influenced by societal biases present in their
training datasets, leading to undesirable differences in how the model responds
when presented with images depicting people of different demographics. In this
work, we propose a training-free debiasing framework for LMMs that intervenes
on the model's representations during text generation by constructing a
steering vector that reduces reference on protected attributes. Our framework
introduces two complementary methods: (1) a dataset-based approach that
constructs a steering vector by contrasting model activations on biased and
neutral inputs, and (2) a novel optimization-based approach designed for
low-resource settings, which constructs the steering vector using a single step
of gradient-based perturbation without requiring additional data. Our
experiments show that these interventions effectively reduce the propensity of
LMMs to generate text related to protected attributes while maintaining
sentiment and fluency. Furthermore, we demonstrate that debiased LMMs achieve
comparable accuracy to their unmodified counterparts on downstream tasks,
indicating that bias mitigation can be achieved without sacrificing model
performance.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.