id
stringlengths 9
16
| submitter
stringlengths 3
64
⌀ | authors
stringlengths 5
6.63k
| title
stringlengths 7
245
| comments
stringlengths 1
482
⌀ | journal-ref
stringlengths 4
382
⌀ | doi
stringlengths 9
151
⌀ | report-no
stringclasses 984
values | categories
stringlengths 5
108
| license
stringclasses 9
values | abstract
stringlengths 83
3.41k
| versions
listlengths 1
20
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-11 00:00:00
| authors_parsed
listlengths 1
427
| prompt
stringlengths 166
3.49k
| label
stringclasses 2
values | prob
float64 0.5
0.98
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2503.09153 | Chaowei Zhang | Chaowei Zhang, Zongling Feng, Zewei Zhang, Jipeng Qiang, Guandong Xu,
and Yun Li | Is LLMs Hallucination Usable? LLM-based Negative Reasoning for Fake News
Detection | 9 pages, 12 figures, conference | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The questionable responses caused by knowledge hallucination may lead to
LLMs' unstable ability in decision-making. However, it has never been
investigated whether the LLMs' hallucination is possibly usable to generate
negative reasoning for facilitating the detection of fake news. This study
proposes a novel supervised self-reinforced reasoning rectification approach -
SR$^3$ that yields both common reasonable reasoning and wrong understandings
(negative reasoning) for news via LLMs reflection for semantic consistency
learning. Upon that, we construct a negative reasoning-based news learning
model called - \emph{NRFE}, which leverages positive or negative news-reasoning
pairs for learning the semantic consistency between them. To avoid the impact
of label-implicated reasoning, we deploy a student model - \emph{NRFE-D} that
only takes news content as input to inspect the performance of our method by
distilling the knowledge from \emph{NRFE}. The experimental results verified on
three popular fake news datasets demonstrate the superiority of our method
compared with three kinds of baselines including prompting on LLMs, fine-tuning
on pre-trained SLMs, and other representative fake news detection methods.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 08:29:59 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zhang",
"Chaowei",
""
],
[
"Feng",
"Zongling",
""
],
[
"Zhang",
"Zewei",
""
],
[
"Qiang",
"Jipeng",
""
],
[
"Xu",
"Guandong",
""
],
[
"Li",
"Yun",
""
]
]
| TITLE: Is LLMs Hallucination Usable? LLM-based Negative Reasoning for Fake News
Detection
ABSTRACT: The questionable responses caused by knowledge hallucination may lead to
LLMs' unstable ability in decision-making. However, it has never been
investigated whether the LLMs' hallucination is possibly usable to generate
negative reasoning for facilitating the detection of fake news. This study
proposes a novel supervised self-reinforced reasoning rectification approach -
SR$^3$ that yields both common reasonable reasoning and wrong understandings
(negative reasoning) for news via LLMs reflection for semantic consistency
learning. Upon that, we construct a negative reasoning-based news learning
model called - \emph{NRFE}, which leverages positive or negative news-reasoning
pairs for learning the semantic consistency between them. To avoid the impact
of label-implicated reasoning, we deploy a student model - \emph{NRFE-D} that
only takes news content as input to inspect the performance of our method by
distilling the knowledge from \emph{NRFE}. The experimental results verified on
three popular fake news datasets demonstrate the superiority of our method
compared with three kinds of baselines including prompting on LLMs, fine-tuning
on pre-trained SLMs, and other representative fake news detection methods.
| no_new_dataset | 0.945551 |
2503.09154 | Yunyang Ge | Chengshu Zhao, Yunyang Ge, Xinhua Cheng, Bin Zhu, Yatian Pang, Bin
Lin, Fan Yang, Feng Gao, Li Yuan | SwapAnyone: Consistent and Realistic Video Synthesis for Swapping Any
Person into Any Video | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video body-swapping aims to replace the body in an existing video with a new
body from arbitrary sources, which has garnered more attention in recent years.
Existing methods treat video body-swapping as a composite of multiple tasks
instead of an independent task and typically rely on various models to achieve
video body-swapping sequentially. However, these methods fail to achieve
end-to-end optimization for the video body-swapping which causes issues such as
variations in luminance among frames, disorganized occlusion relationships, and
the noticeable separation between bodies and background. In this work, we
define video body-swapping as an independent task and propose three critical
consistencies: identity consistency, motion consistency, and environment
consistency. We introduce an end-to-end model named SwapAnyone, treating video
body-swapping as a video inpainting task with reference fidelity and motion
control. To improve the ability to maintain environmental harmony, particularly
luminance harmony in the resulting video, we introduce a novel EnvHarmony
strategy for training our model progressively. Additionally, we provide a
dataset named HumanAction-32K covering various videos about human actions.
Extensive experiments demonstrate that our method achieves State-Of-The-Art
(SOTA) performance among open-source methods while approaching or surpassing
closed-source models across multiple dimensions. All code, model weights, and
the HumanAction-32K dataset will be open-sourced at
https://github.com/PKU-YuanGroup/SwapAnyone.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 08:30:29 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zhao",
"Chengshu",
""
],
[
"Ge",
"Yunyang",
""
],
[
"Cheng",
"Xinhua",
""
],
[
"Zhu",
"Bin",
""
],
[
"Pang",
"Yatian",
""
],
[
"Lin",
"Bin",
""
],
[
"Yang",
"Fan",
""
],
[
"Gao",
"Feng",
""
],
[
"Yuan",
"Li",
""
]
]
| TITLE: SwapAnyone: Consistent and Realistic Video Synthesis for Swapping Any
Person into Any Video
ABSTRACT: Video body-swapping aims to replace the body in an existing video with a new
body from arbitrary sources, which has garnered more attention in recent years.
Existing methods treat video body-swapping as a composite of multiple tasks
instead of an independent task and typically rely on various models to achieve
video body-swapping sequentially. However, these methods fail to achieve
end-to-end optimization for the video body-swapping which causes issues such as
variations in luminance among frames, disorganized occlusion relationships, and
the noticeable separation between bodies and background. In this work, we
define video body-swapping as an independent task and propose three critical
consistencies: identity consistency, motion consistency, and environment
consistency. We introduce an end-to-end model named SwapAnyone, treating video
body-swapping as a video inpainting task with reference fidelity and motion
control. To improve the ability to maintain environmental harmony, particularly
luminance harmony in the resulting video, we introduce a novel EnvHarmony
strategy for training our model progressively. Additionally, we provide a
dataset named HumanAction-32K covering various videos about human actions.
Extensive experiments demonstrate that our method achieves State-Of-The-Art
(SOTA) performance among open-source methods while approaching or surpassing
closed-source models across multiple dimensions. All code, model weights, and
the HumanAction-32K dataset will be open-sourced at
https://github.com/PKU-YuanGroup/SwapAnyone.
| new_dataset | 0.964456 |
2503.09159 | Andrej Tschalzev | Andrej Tschalzev, Lennart Purucker, Stefan L\"udtke, Frank Hutter,
Christian Bartelt, Heiner Stuckenschmidt | Unreflected Use of Tabular Data Repositories Can Undermine Research
Quality | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Data repositories have accumulated a large number of tabular datasets from
various domains. Machine Learning researchers are actively using these datasets
to evaluate novel approaches. Consequently, data repositories have an important
standing in tabular data research. They not only host datasets but also provide
information on how to use them in supervised learning tasks. In this paper, we
argue that, despite great achievements in usability, the unreflected usage of
datasets from data repositories may have led to reduced research quality and
scientific rigor. We present examples from prominent recent studies that
illustrate the problematic use of datasets from OpenML, a large data repository
for tabular data. Our illustrations help users of data repositories avoid
falling into the traps of (1) using suboptimal model selection strategies, (2)
overlooking strong baselines, and (3) inappropriate preprocessing. In response,
we discuss possible solutions for how data repositories can prevent the
inappropriate use of datasets and become the cornerstones for improved overall
quality of empirical research studies.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 08:41:49 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Tschalzev",
"Andrej",
""
],
[
"Purucker",
"Lennart",
""
],
[
"Lüdtke",
"Stefan",
""
],
[
"Hutter",
"Frank",
""
],
[
"Bartelt",
"Christian",
""
],
[
"Stuckenschmidt",
"Heiner",
""
]
]
| TITLE: Unreflected Use of Tabular Data Repositories Can Undermine Research
Quality
ABSTRACT: Data repositories have accumulated a large number of tabular datasets from
various domains. Machine Learning researchers are actively using these datasets
to evaluate novel approaches. Consequently, data repositories have an important
standing in tabular data research. They not only host datasets but also provide
information on how to use them in supervised learning tasks. In this paper, we
argue that, despite great achievements in usability, the unreflected usage of
datasets from data repositories may have led to reduced research quality and
scientific rigor. We present examples from prominent recent studies that
illustrate the problematic use of datasets from OpenML, a large data repository
for tabular data. Our illustrations help users of data repositories avoid
falling into the traps of (1) using suboptimal model selection strategies, (2)
overlooking strong baselines, and (3) inappropriate preprocessing. In response,
we discuss possible solutions for how data repositories can prevent the
inappropriate use of datasets and become the cornerstones for improved overall
quality of empirical research studies.
| no_new_dataset | 0.952706 |
2503.09170 | Aman Prakash | Aman Prakash, Nikumani Choudhury, Anakhi Hazarika, Alekhya Gorrela | Effective Feature Selection for Predicting Spreading Factor with ML in
Large LoRaWAN-based Mobile IoT Networks | Accepted at 31st National Conference on Communications | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | LoRaWAN is a low-power long-range protocol that enables reliable and robust
communication. This paper addresses the challenge of predicting the spreading
factor (SF) in LoRaWAN networks using machine learning (ML) techniques. Optimal
SF allocation is crucial for optimizing data transmission in IoT-enabled mobile
devices, yet it remains a challenging task due to the fluctuation in
environment and network conditions. We evaluated ML model performance across a
large publicly available dataset to explore the best feature across key LoRaWAN
features such as RSSI, SNR, frequency, distance between end devices and
gateways, and antenna height of the end device, further, we also experimented
with 31 different combinations possible for 5 features. We trained and
evaluated the model using k-nearest neighbors (k-NN), Decision Tree Classifier
(DTC), Random Forest (RF), and Multinomial Logistic Regression (MLR)
algorithms. The combination of RSSI and SNR was identified as the best feature
set. The finding of this paper provides valuable information for reducing the
overall cost of dataset collection for ML model training and extending the
battery life of LoRaWAN devices. This work contributes to a more reliable
LoRaWAN system by understanding the importance of specific feature sets for
optimized SF allocation.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 08:58:28 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Prakash",
"Aman",
""
],
[
"Choudhury",
"Nikumani",
""
],
[
"Hazarika",
"Anakhi",
""
],
[
"Gorrela",
"Alekhya",
""
]
]
| TITLE: Effective Feature Selection for Predicting Spreading Factor with ML in
Large LoRaWAN-based Mobile IoT Networks
ABSTRACT: LoRaWAN is a low-power long-range protocol that enables reliable and robust
communication. This paper addresses the challenge of predicting the spreading
factor (SF) in LoRaWAN networks using machine learning (ML) techniques. Optimal
SF allocation is crucial for optimizing data transmission in IoT-enabled mobile
devices, yet it remains a challenging task due to the fluctuation in
environment and network conditions. We evaluated ML model performance across a
large publicly available dataset to explore the best feature across key LoRaWAN
features such as RSSI, SNR, frequency, distance between end devices and
gateways, and antenna height of the end device, further, we also experimented
with 31 different combinations possible for 5 features. We trained and
evaluated the model using k-nearest neighbors (k-NN), Decision Tree Classifier
(DTC), Random Forest (RF), and Multinomial Logistic Regression (MLR)
algorithms. The combination of RSSI and SNR was identified as the best feature
set. The finding of this paper provides valuable information for reducing the
overall cost of dataset collection for ML model training and extending the
battery life of LoRaWAN devices. This work contributes to a more reliable
LoRaWAN system by understanding the importance of specific feature sets for
optimized SF allocation.
| no_new_dataset | 0.953319 |
2503.09181 | Katsumi Takahashi | Katsumi Takahashi, Koh Takeuchi, Hisashi Kashima | Dynamic Feature Selection from Variable Feature Sets Using Features of
Features | null | null | null | null | cs.LG cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning models usually assume that a set of feature values used to
obtain an output is fixed in advance. However, in many real-world problems, a
cost is associated with measuring these features. To address the issue of
reducing measurement costs, various methods have been proposed to dynamically
select which features to measure, but existing methods assume that the set of
measurable features remains constant, which makes them unsuitable for cases
where the set of measurable features varies from instance to instance. To
overcome this limitation, we define a new problem setting for Dynamic Feature
Selection (DFS) with variable feature sets and propose a deep learning method
that utilizes prior information about each feature, referred to as ''features
of features''. Experimental results on several datasets demonstrate that the
proposed method effectively selects features based on the prior information,
even when the set of measurable features changes from instance to instance.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:13:21 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Takahashi",
"Katsumi",
""
],
[
"Takeuchi",
"Koh",
""
],
[
"Kashima",
"Hisashi",
""
]
]
| TITLE: Dynamic Feature Selection from Variable Feature Sets Using Features of
Features
ABSTRACT: Machine learning models usually assume that a set of feature values used to
obtain an output is fixed in advance. However, in many real-world problems, a
cost is associated with measuring these features. To address the issue of
reducing measurement costs, various methods have been proposed to dynamically
select which features to measure, but existing methods assume that the set of
measurable features remains constant, which makes them unsuitable for cases
where the set of measurable features varies from instance to instance. To
overcome this limitation, we define a new problem setting for Dynamic Feature
Selection (DFS) with variable feature sets and propose a deep learning method
that utilizes prior information about each feature, referred to as ''features
of features''. Experimental results on several datasets demonstrate that the
proposed method effectively selects features based on the prior information,
even when the set of measurable features changes from instance to instance.
| no_new_dataset | 0.945901 |
2503.09185 | Yuanyang Zhang | Yuanyang Zhang, Yijie Lin, Weiqing Yan, Li Yao, Xinhang Wan, Guangyuan
Li, Chao Zhang, Guanzhou Ke, Jie Xu | Incomplete Multi-view Clustering via Diffusion Contrastive Generation | null | AAAI 2025 | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Incomplete multi-view clustering (IMVC) has garnered increasing attention in
recent years due to the common issue of missing data in multi-view datasets.
The primary approach to address this challenge involves recovering the missing
views before applying conventional multi-view clustering methods. Although
imputation-based IMVC methods have achieved significant improvements, they
still encounter notable limitations: 1) heavy reliance on paired data for
training the data recovery module, which is impractical in real scenarios with
high missing data rates; 2) the generated data often lacks diversity and
discriminability, resulting in suboptimal clustering results. To address these
shortcomings, we propose a novel IMVC method called Diffusion Contrastive
Generation (DCG). Motivated by the consistency between the diffusion and
clustering processes, DCG learns the distribution characteristics to enhance
clustering by applying forward diffusion and reverse denoising processes to
intra-view data. By performing contrastive learning on a limited set of paired
multi-view samples, DCG can align the generated views with the real views,
facilitating accurate recovery of views across arbitrary missing view
scenarios. Additionally, DCG integrates instance-level and category-level
interactive learning to exploit the consistent and complementary information
available in multi-view data, achieving robust and end-to-end clustering.
Extensive experiments demonstrate that our method outperforms state-of-the-art
approaches.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:27:25 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zhang",
"Yuanyang",
""
],
[
"Lin",
"Yijie",
""
],
[
"Yan",
"Weiqing",
""
],
[
"Yao",
"Li",
""
],
[
"Wan",
"Xinhang",
""
],
[
"Li",
"Guangyuan",
""
],
[
"Zhang",
"Chao",
""
],
[
"Ke",
"Guanzhou",
""
],
[
"Xu",
"Jie",
""
]
]
| TITLE: Incomplete Multi-view Clustering via Diffusion Contrastive Generation
ABSTRACT: Incomplete multi-view clustering (IMVC) has garnered increasing attention in
recent years due to the common issue of missing data in multi-view datasets.
The primary approach to address this challenge involves recovering the missing
views before applying conventional multi-view clustering methods. Although
imputation-based IMVC methods have achieved significant improvements, they
still encounter notable limitations: 1) heavy reliance on paired data for
training the data recovery module, which is impractical in real scenarios with
high missing data rates; 2) the generated data often lacks diversity and
discriminability, resulting in suboptimal clustering results. To address these
shortcomings, we propose a novel IMVC method called Diffusion Contrastive
Generation (DCG). Motivated by the consistency between the diffusion and
clustering processes, DCG learns the distribution characteristics to enhance
clustering by applying forward diffusion and reverse denoising processes to
intra-view data. By performing contrastive learning on a limited set of paired
multi-view samples, DCG can align the generated views with the real views,
facilitating accurate recovery of views across arbitrary missing view
scenarios. Additionally, DCG integrates instance-level and category-level
interactive learning to exploit the consistent and complementary information
available in multi-view data, achieving robust and end-to-end clustering.
Extensive experiments demonstrate that our method outperforms state-of-the-art
approaches.
| no_new_dataset | 0.948106 |
2503.09186 | Jian-Jian Jiang | Jian-Jian Jiang, Xiao-Ming Wu, Yi-Xiang He, Ling-An Zeng, Yi-Lin Wei,
Dandan Zhang, Wei-Shi Zheng | Rethinking Bimanual Robotic Manipulation: Learning with Decoupled
Interaction Framework | 14 pages, 8 figures | null | null | null | cs.RO cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bimanual robotic manipulation is an emerging and critical topic in the
robotics community. Previous works primarily rely on integrated control models
that take the perceptions and states of both arms as inputs to directly predict
their actions. However, we think bimanual manipulation involves not only
coordinated tasks but also various uncoordinated tasks that do not require
explicit cooperation during execution, such as grasping objects with the
closest hand, which integrated control frameworks ignore to consider due to
their enforced cooperation in the early inputs. In this paper, we propose a
novel decoupled interaction framework that considers the characteristics of
different tasks in bimanual manipulation. The key insight of our framework is
to assign an independent model to each arm to enhance the learning of
uncoordinated tasks, while introducing a selective interaction module that
adaptively learns weights from its own arm to improve the learning of
coordinated tasks. Extensive experiments on seven tasks in the RoboTwin dataset
demonstrate that: (1) Our framework achieves outstanding performance, with a
23.5% boost over the SOTA method. (2) Our framework is flexible and can be
seamlessly integrated into existing methods. (3) Our framework can be
effectively extended to multi-agent manipulation tasks, achieving a 28% boost
over the integrated control SOTA. (4) The performance boost stems from the
decoupled design itself, surpassing the SOTA by 16.5% in success rate with only
1/6 of the model size.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:28:41 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Jiang",
"Jian-Jian",
""
],
[
"Wu",
"Xiao-Ming",
""
],
[
"He",
"Yi-Xiang",
""
],
[
"Zeng",
"Ling-An",
""
],
[
"Wei",
"Yi-Lin",
""
],
[
"Zhang",
"Dandan",
""
],
[
"Zheng",
"Wei-Shi",
""
]
]
| TITLE: Rethinking Bimanual Robotic Manipulation: Learning with Decoupled
Interaction Framework
ABSTRACT: Bimanual robotic manipulation is an emerging and critical topic in the
robotics community. Previous works primarily rely on integrated control models
that take the perceptions and states of both arms as inputs to directly predict
their actions. However, we think bimanual manipulation involves not only
coordinated tasks but also various uncoordinated tasks that do not require
explicit cooperation during execution, such as grasping objects with the
closest hand, which integrated control frameworks ignore to consider due to
their enforced cooperation in the early inputs. In this paper, we propose a
novel decoupled interaction framework that considers the characteristics of
different tasks in bimanual manipulation. The key insight of our framework is
to assign an independent model to each arm to enhance the learning of
uncoordinated tasks, while introducing a selective interaction module that
adaptively learns weights from its own arm to improve the learning of
coordinated tasks. Extensive experiments on seven tasks in the RoboTwin dataset
demonstrate that: (1) Our framework achieves outstanding performance, with a
23.5% boost over the SOTA method. (2) Our framework is flexible and can be
seamlessly integrated into existing methods. (3) Our framework can be
effectively extended to multi-agent manipulation tasks, achieving a 28% boost
over the integrated control SOTA. (4) The performance boost stems from the
decoupled design itself, surpassing the SOTA by 16.5% in success rate with only
1/6 of the model size.
| no_new_dataset | 0.9463 |
2503.09187 | Qipeng Mei | Qipeng Mei, Dimitri Bulatov and Dorota Iwaszczuk | Polygonizing Roof Segments from High-Resolution Aerial Images Using
Yolov8-Based Edge Detection | 12 pages, 6 figures, conference paper (VISAPP 2025, part of the 20th
International Joint Conference on Computer Vision, Imaging, and Computer
Graphics Theory and Applications) | null | 10.5220/0013130400003912 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This study presents a novel approach for roof detail extraction and
vectorization using remote sensing images. Unlike previous
geometric-primitive-based methods that rely on the detection of corners, our
method focuses on edge detection as the primary mechanism for roof
reconstruction, while utilizing geometric relationships to define corners and
faces. We adapt the YOLOv8 OBB model, originally designed for rotated object
detection, to extract roof edges effectively. Our method demonstrates
robustness against noise and occlusion, leading to precise vectorized
representations of building roofs. Experiments conducted on the SGA and
Melville datasets highlight the method's effectiveness. At the raster level,
our model outperforms the state-of-the-art foundation segmentation model (SAM),
achieving a mIoU between 0.85 and 1 for most samples and an ovIoU close to
0.97. At the vector level, evaluation using the Hausdorff distance, PolyS
metric, and our raster-vector-metric demonstrates significant improvements
after polygonization, with a close approximation to the reference data. The
method successfully handles diverse roof structures and refines edge gaps, even
on complex roof structures of new, excluded from training datasets. Our
findings underscore the potential of this approach to address challenges in
automatic roof structure vectorization, supporting various applications such as
urban terrain reconstruction.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:29:10 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Mei",
"Qipeng",
""
],
[
"Bulatov",
"Dimitri",
""
],
[
"Iwaszczuk",
"Dorota",
""
]
]
| TITLE: Polygonizing Roof Segments from High-Resolution Aerial Images Using
Yolov8-Based Edge Detection
ABSTRACT: This study presents a novel approach for roof detail extraction and
vectorization using remote sensing images. Unlike previous
geometric-primitive-based methods that rely on the detection of corners, our
method focuses on edge detection as the primary mechanism for roof
reconstruction, while utilizing geometric relationships to define corners and
faces. We adapt the YOLOv8 OBB model, originally designed for rotated object
detection, to extract roof edges effectively. Our method demonstrates
robustness against noise and occlusion, leading to precise vectorized
representations of building roofs. Experiments conducted on the SGA and
Melville datasets highlight the method's effectiveness. At the raster level,
our model outperforms the state-of-the-art foundation segmentation model (SAM),
achieving a mIoU between 0.85 and 1 for most samples and an ovIoU close to
0.97. At the vector level, evaluation using the Hausdorff distance, PolyS
metric, and our raster-vector-metric demonstrates significant improvements
after polygonization, with a close approximation to the reference data. The
method successfully handles diverse roof structures and refines edge gaps, even
on complex roof structures of new, excluded from training datasets. Our
findings underscore the potential of this approach to address challenges in
automatic roof structure vectorization, supporting various applications such as
urban terrain reconstruction.
| no_new_dataset | 0.949995 |
2503.09191 | Juana Valeria Hurtado | Juana Valeria Hurtado, Sajad Marvi, Rohit Mohan, and Abhinav Valada | Learning Appearance and Motion Cues for Panoptic Tracking | null | null | null | null | cs.CV cs.RO | http://creativecommons.org/licenses/by/4.0/ | Panoptic tracking enables pixel-level scene interpretation of videos by
integrating instance tracking in panoptic segmentation. This provides robots
with a spatio-temporal understanding of the environment, an essential attribute
for their operation in dynamic environments. In this paper, we propose a novel
approach for panoptic tracking that simultaneously captures general semantic
information and instance-specific appearance and motion features. Unlike
existing methods that overlook dynamic scene attributes, our approach leverages
both appearance and motion cues through dedicated network heads. These
interconnected heads employ multi-scale deformable convolutions that reason
about scene motion offsets with semantic context and motion-enhanced appearance
features to learn tracking embeddings. Furthermore, we introduce a novel
two-step fusion module that integrates the outputs from both heads by first
matching instances from the current time step with propagated instances from
previous time steps and subsequently refines associations using motion-enhanced
appearance embeddings, improving robustness in challenging scenarios. Extensive
evaluations of our proposed \netname model on two benchmark datasets
demonstrate that it achieves state-of-the-art performance in panoptic tracking
accuracy, surpassing prior methods in maintaining object identities over time.
To facilitate future research, we make the code available at
http://panoptictracking.cs.uni-freiburg.de
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:32:29 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Hurtado",
"Juana Valeria",
""
],
[
"Marvi",
"Sajad",
""
],
[
"Mohan",
"Rohit",
""
],
[
"Valada",
"Abhinav",
""
]
]
| TITLE: Learning Appearance and Motion Cues for Panoptic Tracking
ABSTRACT: Panoptic tracking enables pixel-level scene interpretation of videos by
integrating instance tracking in panoptic segmentation. This provides robots
with a spatio-temporal understanding of the environment, an essential attribute
for their operation in dynamic environments. In this paper, we propose a novel
approach for panoptic tracking that simultaneously captures general semantic
information and instance-specific appearance and motion features. Unlike
existing methods that overlook dynamic scene attributes, our approach leverages
both appearance and motion cues through dedicated network heads. These
interconnected heads employ multi-scale deformable convolutions that reason
about scene motion offsets with semantic context and motion-enhanced appearance
features to learn tracking embeddings. Furthermore, we introduce a novel
two-step fusion module that integrates the outputs from both heads by first
matching instances from the current time step with propagated instances from
previous time steps and subsequently refines associations using motion-enhanced
appearance embeddings, improving robustness in challenging scenarios. Extensive
evaluations of our proposed \netname model on two benchmark datasets
demonstrate that it achieves state-of-the-art performance in panoptic tracking
accuracy, surpassing prior methods in maintaining object identities over time.
To facilitate future research, we make the code available at
http://panoptictracking.cs.uni-freiburg.de
| no_new_dataset | 0.950915 |
2503.09196 | Sanggyu Chong | Federico Grasselli, Sanggyu Chong, Venkat Kapil, Silvia Bonfanti and
Kevin Rossi | Uncertainty in the era of machine learning for atomistic modeling | 21 pages, 8 figures | null | null | null | physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | The widespread adoption of machine learning surrogate models has
significantly improved the scale and complexity of systems and processes that
can be explored accurately and efficiently using atomistic modeling. However,
the inherently data-driven nature of machine learning models introduces
uncertainties that must be quantified, understood, and effectively managed to
ensure reliable predictions and conclusions. Building upon these premises, in
this Perspective, we first overview state-of-the-art uncertainty estimation
methods, from Bayesian frameworks to ensembling techniques, and discuss their
application in atomistic modeling. We then examine the interplay between model
accuracy, uncertainty, training dataset composition, data acquisition
strategies, model transferability, and robustness. In doing so, we synthesize
insights from the existing literature and highlight areas of ongoing debate.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:39:18 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Grasselli",
"Federico",
""
],
[
"Chong",
"Sanggyu",
""
],
[
"Kapil",
"Venkat",
""
],
[
"Bonfanti",
"Silvia",
""
],
[
"Rossi",
"Kevin",
""
]
]
| TITLE: Uncertainty in the era of machine learning for atomistic modeling
ABSTRACT: The widespread adoption of machine learning surrogate models has
significantly improved the scale and complexity of systems and processes that
can be explored accurately and efficiently using atomistic modeling. However,
the inherently data-driven nature of machine learning models introduces
uncertainties that must be quantified, understood, and effectively managed to
ensure reliable predictions and conclusions. Building upon these premises, in
this Perspective, we first overview state-of-the-art uncertainty estimation
methods, from Bayesian frameworks to ensembling techniques, and discuss their
application in atomistic modeling. We then examine the interplay between model
accuracy, uncertainty, training dataset composition, data acquisition
strategies, model transferability, and robustness. In doing so, we synthesize
insights from the existing literature and highlight areas of ongoing debate.
| no_new_dataset | 0.950088 |
2503.09197 | Zicheng Zhang | Zicheng Zhang, Haoning Wu, Ziheng Jia, Weisi Lin, Guangtao Zhai | Teaching LMMs for Image Quality Scoring and Interpreting | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Image quality scoring and interpreting are two fundamental components of
Image Quality Assessment (IQA). The former quantifies image quality, while the
latter enables descriptive question answering about image quality.
Traditionally, these two tasks have been addressed independently. However, from
the perspective of the Human Visual System (HVS) and the Perception-Decision
Integration Model, they are inherently interconnected: interpreting serves as
the foundation for scoring, while scoring provides an abstract summary of
interpreting. Thus, unifying these capabilities within a single model is both
intuitive and logically coherent. In this paper, we propose Q-SiT (Quality
Scoring and Interpreting joint Teaching), a unified framework that enables
large multimodal models (LMMs) to learn both image quality scoring and
interpreting simultaneously. We achieve this by transforming conventional IQA
datasets into learnable question-answering datasets and incorporating
human-annotated quality interpreting data for training. Furthermore, we
introduce an efficient scoring & interpreting balance strategy, which first
determines the optimal data mix ratio on lightweight LMMs and then maps this
ratio to primary LMMs for fine-tuning adjustment. This strategy not only
mitigates task interference and enhances cross-task knowledge transfer but also
significantly reduces computational costs compared to direct optimization on
full-scale LMMs. With this joint learning framework and corresponding training
strategy, we develop Q-SiT, the first model capable of simultaneously
performing image quality scoring and interpreting tasks, along with its
lightweight variant, Q-SiT-mini. Experimental results demonstrate that Q-SiT
achieves strong performance in both tasks with superior generalization IQA
abilities.Project page at https://github.com/Q-Future/Q-SiT.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:39:33 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zhang",
"Zicheng",
""
],
[
"Wu",
"Haoning",
""
],
[
"Jia",
"Ziheng",
""
],
[
"Lin",
"Weisi",
""
],
[
"Zhai",
"Guangtao",
""
]
]
| TITLE: Teaching LMMs for Image Quality Scoring and Interpreting
ABSTRACT: Image quality scoring and interpreting are two fundamental components of
Image Quality Assessment (IQA). The former quantifies image quality, while the
latter enables descriptive question answering about image quality.
Traditionally, these two tasks have been addressed independently. However, from
the perspective of the Human Visual System (HVS) and the Perception-Decision
Integration Model, they are inherently interconnected: interpreting serves as
the foundation for scoring, while scoring provides an abstract summary of
interpreting. Thus, unifying these capabilities within a single model is both
intuitive and logically coherent. In this paper, we propose Q-SiT (Quality
Scoring and Interpreting joint Teaching), a unified framework that enables
large multimodal models (LMMs) to learn both image quality scoring and
interpreting simultaneously. We achieve this by transforming conventional IQA
datasets into learnable question-answering datasets and incorporating
human-annotated quality interpreting data for training. Furthermore, we
introduce an efficient scoring & interpreting balance strategy, which first
determines the optimal data mix ratio on lightweight LMMs and then maps this
ratio to primary LMMs for fine-tuning adjustment. This strategy not only
mitigates task interference and enhances cross-task knowledge transfer but also
significantly reduces computational costs compared to direct optimization on
full-scale LMMs. With this joint learning framework and corresponding training
strategy, we develop Q-SiT, the first model capable of simultaneously
performing image quality scoring and interpreting tasks, along with its
lightweight variant, Q-SiT-mini. Experimental results demonstrate that Q-SiT
achieves strong performance in both tasks with superior generalization IQA
abilities.Project page at https://github.com/Q-Future/Q-SiT.
| no_new_dataset | 0.954351 |
2503.09200 | Chichun Zhou | Lei Liu, Yuchao Lu, Ling An, Huajie Liang, Chichun Zhou, Zhenyu Zhang | Time-EAPCR: A Deep Learning-Based Novel Approach for Anomaly Detection
Applied to the Environmental Field | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As human activities intensify, environmental systems such as aquatic
ecosystems and water treatment systems face increasingly complex pressures,
impacting ecological balance, public health, and sustainable development,
making intelligent anomaly monitoring essential. However, traditional
monitoring methods suffer from delayed responses, insufficient data processing
capabilities, and weak generalisation, making them unsuitable for complex
environmental monitoring needs.In recent years, machine learning has been
widely applied to anomaly detection, but the multi-dimensional features and
spatiotemporal dynamics of environmental ecological data, especially the
long-term dependencies and strong variability in the time dimension, limit the
effectiveness of traditional methods.Deep learning, with its ability to
automatically learn features, captures complex nonlinear relationships,
improving detection performance. However, its application in environmental
monitoring is still in its early stages and requires further exploration.This
paper introduces a new deep learning method, Time-EAPCR
(Time-Embedding-Attention-Permutated CNN-Residual), and applies it to
environmental science. The method uncovers feature correlations, captures
temporal evolution patterns, and enables precise anomaly detection in
environmental systems.We validated Time-EAPCR's high accuracy and robustness
across four publicly available environmental datasets. Experimental results
show that the method efficiently handles multi-source data, improves detection
accuracy, and excels across various scenarios with strong adaptability and
generalisation. Additionally, a real-world river monitoring dataset confirmed
the feasibility of its deployment, providing reliable technical support for
environmental monitoring.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 09:44:15 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Liu",
"Lei",
""
],
[
"Lu",
"Yuchao",
""
],
[
"An",
"Ling",
""
],
[
"Liang",
"Huajie",
""
],
[
"Zhou",
"Chichun",
""
],
[
"Zhang",
"Zhenyu",
""
]
]
| TITLE: Time-EAPCR: A Deep Learning-Based Novel Approach for Anomaly Detection
Applied to the Environmental Field
ABSTRACT: As human activities intensify, environmental systems such as aquatic
ecosystems and water treatment systems face increasingly complex pressures,
impacting ecological balance, public health, and sustainable development,
making intelligent anomaly monitoring essential. However, traditional
monitoring methods suffer from delayed responses, insufficient data processing
capabilities, and weak generalisation, making them unsuitable for complex
environmental monitoring needs.In recent years, machine learning has been
widely applied to anomaly detection, but the multi-dimensional features and
spatiotemporal dynamics of environmental ecological data, especially the
long-term dependencies and strong variability in the time dimension, limit the
effectiveness of traditional methods.Deep learning, with its ability to
automatically learn features, captures complex nonlinear relationships,
improving detection performance. However, its application in environmental
monitoring is still in its early stages and requires further exploration.This
paper introduces a new deep learning method, Time-EAPCR
(Time-Embedding-Attention-Permutated CNN-Residual), and applies it to
environmental science. The method uncovers feature correlations, captures
temporal evolution patterns, and enables precise anomaly detection in
environmental systems.We validated Time-EAPCR's high accuracy and robustness
across four publicly available environmental datasets. Experimental results
show that the method efficiently handles multi-source data, improves detection
accuracy, and excels across various scenarios with strong adaptability and
generalisation. Additionally, a real-world river monitoring dataset confirmed
the feasibility of its deployment, providing reliable technical support for
environmental monitoring.
| no_new_dataset | 0.941223 |
2503.09217 | Fengjie Li | Fengjie Li, Jiajun Jiang, Jiajun Sun, Hongyu Zhang | Evaluating the Generalizability of LLMs in Automated Program Repair | 5 pages, 1 figure, to be published in ICSE2025-NIER | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | LLM-based automated program repair methods have attracted significant
attention for their state-of-the-art performance. However, they were primarily
evaluated on a few well known datasets like Defects4J, raising questions about
their effectiveness on new datasets. In this study, we evaluate 11
top-performing LLMs on DEFECTS4J-TRANS, a new dataset derived from transforming
Defects4J while maintaining the original semantics. Results from experiments on
both Defects4J and DEFECTS4J-TRANS show that all studied LLMs have limited
generalizability in APR tasks, with the average number of correct and plausible
patches decreasing by 49.48% and 42.90%, respectively, on DEFECTS4J-TRANS.
Further investigation into incorporating additional repair-relevant information
in repair prompts reveals that, although this information significantly
enhances the LLMs' capabilities (increasing the number of correct and plausible
patches by up to 136.67% and 121.82%, respectively), performance still falls
short of their original results. This indicates that prompt engineering alone
is insufficient to substantially enhance LLMs' repair capabilities. Based on
our study, we also offer several recommendations for future research.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:03:58 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Li",
"Fengjie",
""
],
[
"Jiang",
"Jiajun",
""
],
[
"Sun",
"Jiajun",
""
],
[
"Zhang",
"Hongyu",
""
]
]
| TITLE: Evaluating the Generalizability of LLMs in Automated Program Repair
ABSTRACT: LLM-based automated program repair methods have attracted significant
attention for their state-of-the-art performance. However, they were primarily
evaluated on a few well known datasets like Defects4J, raising questions about
their effectiveness on new datasets. In this study, we evaluate 11
top-performing LLMs on DEFECTS4J-TRANS, a new dataset derived from transforming
Defects4J while maintaining the original semantics. Results from experiments on
both Defects4J and DEFECTS4J-TRANS show that all studied LLMs have limited
generalizability in APR tasks, with the average number of correct and plausible
patches decreasing by 49.48% and 42.90%, respectively, on DEFECTS4J-TRANS.
Further investigation into incorporating additional repair-relevant information
in repair prompts reveals that, although this information significantly
enhances the LLMs' capabilities (increasing the number of correct and plausible
patches by up to 136.67% and 121.82%, respectively), performance still falls
short of their original results. This indicates that prompt engineering alone
is insufficient to substantially enhance LLMs' repair capabilities. Based on
our study, we also offer several recommendations for future research.
| no_new_dataset | 0.647993 |
2503.09218 | Jie He | Jie He, Simon Yu, Deyi Xiong, V\'ictor Guti\'errez-Basulto, Jeff Z.
Pan | N2C2: Nearest Neighbor Enhanced Confidence Calibration for Cross-Lingual
In-Context Learning | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recent advancements of in-context learning (ICL) show language models can
significantly improve their performance when demonstrations are provided.
However, little attention has been paid to model calibration and prediction
confidence of ICL in cross-lingual scenarios. To bridge this gap, we conduct a
thorough analysis of ICL for cross-lingual sentiment classification. Our
findings suggest that ICL performs poorly in cross-lingual scenarios,
exhibiting low accuracy and presenting high calibration errors. In response, we
propose a novel approach, N2C2, which employs a -nearest neighbors augmented
classifier for prediction confidence calibration. N2C2 narrows the prediction
gap by leveraging a datastore of cached few-shot instances. Specifically, N2C2
integrates the predictions from the datastore and incorporates confidence-aware
distribution, semantically consistent retrieval representation, and adaptive
neighbor combination modules to effectively utilize the limited number of
supporting instances. Evaluation on two multilingual sentiment classification
datasets demonstrates that N2C2 outperforms traditional ICL. It surpasses fine
tuning, prompt tuning and recent state-of-the-art methods in terms of accuracy
and calibration errors.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:05:05 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"He",
"Jie",
""
],
[
"Yu",
"Simon",
""
],
[
"Xiong",
"Deyi",
""
],
[
"Gutiérrez-Basulto",
"Víctor",
""
],
[
"Pan",
"Jeff Z.",
""
]
]
| TITLE: N2C2: Nearest Neighbor Enhanced Confidence Calibration for Cross-Lingual
In-Context Learning
ABSTRACT: Recent advancements of in-context learning (ICL) show language models can
significantly improve their performance when demonstrations are provided.
However, little attention has been paid to model calibration and prediction
confidence of ICL in cross-lingual scenarios. To bridge this gap, we conduct a
thorough analysis of ICL for cross-lingual sentiment classification. Our
findings suggest that ICL performs poorly in cross-lingual scenarios,
exhibiting low accuracy and presenting high calibration errors. In response, we
propose a novel approach, N2C2, which employs a -nearest neighbors augmented
classifier for prediction confidence calibration. N2C2 narrows the prediction
gap by leveraging a datastore of cached few-shot instances. Specifically, N2C2
integrates the predictions from the datastore and incorporates confidence-aware
distribution, semantically consistent retrieval representation, and adaptive
neighbor combination modules to effectively utilize the limited number of
supporting instances. Evaluation on two multilingual sentiment classification
datasets demonstrates that N2C2 outperforms traditional ICL. It surpasses fine
tuning, prompt tuning and recent state-of-the-art methods in terms of accuracy
and calibration errors.
| no_new_dataset | 0.943608 |
2503.09219 | Runzhe Zhan | Xinyi Yang, Runzhe Zhan, Derek F. Wong, Shu Yang, Junchao Wu, Lidia S.
Chao | Rethinking Prompt-based Debiasing in Large Language Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Investigating bias in large language models (LLMs) is crucial for developing
trustworthy AI. While prompt-based through prompt engineering is common, its
effectiveness relies on the assumption that models inherently understand
biases. Our study systematically analyzed this assumption using the BBQ and
StereoSet benchmarks on both open-source models as well as commercial GPT
model. Experimental results indicate that prompt-based is often superficial;
for instance, the Llama2-7B-Chat model misclassified over 90% of unbiased
content as biased, despite achieving high accuracy in identifying bias issues
on the BBQ dataset. Additionally, specific evaluation and question settings in
bias benchmarks often lead LLMs to choose "evasive answers", disregarding the
core of the question and the relevance of the response to the context.
Moreover, the apparent success of previous methods may stem from flawed
evaluation metrics. Our research highlights a potential "false prosperity" in
prompt-base efforts and emphasizes the need to rethink bias metrics to ensure
truly trustworthy AI.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:06:03 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Yang",
"Xinyi",
""
],
[
"Zhan",
"Runzhe",
""
],
[
"Wong",
"Derek F.",
""
],
[
"Yang",
"Shu",
""
],
[
"Wu",
"Junchao",
""
],
[
"Chao",
"Lidia S.",
""
]
]
| TITLE: Rethinking Prompt-based Debiasing in Large Language Models
ABSTRACT: Investigating bias in large language models (LLMs) is crucial for developing
trustworthy AI. While prompt-based through prompt engineering is common, its
effectiveness relies on the assumption that models inherently understand
biases. Our study systematically analyzed this assumption using the BBQ and
StereoSet benchmarks on both open-source models as well as commercial GPT
model. Experimental results indicate that prompt-based is often superficial;
for instance, the Llama2-7B-Chat model misclassified over 90% of unbiased
content as biased, despite achieving high accuracy in identifying bias issues
on the BBQ dataset. Additionally, specific evaluation and question settings in
bias benchmarks often lead LLMs to choose "evasive answers", disregarding the
core of the question and the relevance of the response to the context.
Moreover, the apparent success of previous methods may stem from flawed
evaluation metrics. Our research highlights a potential "false prosperity" in
prompt-base efforts and emphasizes the need to rethink bias metrics to ensure
truly trustworthy AI.
| no_new_dataset | 0.947381 |
2503.09221 | Hannah Kniesel | Hannah Kniesel, Pedro Hermosilla, Timo Ropinski | Active Learning Inspired ControlNet Guidance for Augmenting Semantic
Segmentation Datasets | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in conditional image generation from diffusion models have
shown great potential in achieving impressive image quality while preserving
the constraints introduced by the user. In particular, ControlNet enables
precise alignment between ground truth segmentation masks and the generated
image content, allowing the enhancement of training datasets in segmentation
tasks. This raises a key question: Can ControlNet additionally be guided to
generate the most informative synthetic samples for a specific task? Inspired
by active learning, where the most informative real-world samples are selected
based on sample difficulty or model uncertainty, we propose the first approach
to integrate active learning-based selection metrics into the backward
diffusion process for sample generation. Specifically, we explore uncertainty,
query by committee, and expected model change, which are commonly used in
active learning, and demonstrate their application for guiding the sample
generation process through gradient approximation. Our method is training-free,
modifying only the backward diffusion process, allowing it to be used on any
pretrained ControlNet. Using this process, we show that segmentation models
trained with guided synthetic data outperform those trained on non-guided
synthetic data. Our work underscores the need for advanced control mechanisms
for diffusion-based models, which are not only aligned with image content but
additionally downstream task performance, highlighting the true potential of
synthetic data generation.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:09:27 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Kniesel",
"Hannah",
""
],
[
"Hermosilla",
"Pedro",
""
],
[
"Ropinski",
"Timo",
""
]
]
| TITLE: Active Learning Inspired ControlNet Guidance for Augmenting Semantic
Segmentation Datasets
ABSTRACT: Recent advances in conditional image generation from diffusion models have
shown great potential in achieving impressive image quality while preserving
the constraints introduced by the user. In particular, ControlNet enables
precise alignment between ground truth segmentation masks and the generated
image content, allowing the enhancement of training datasets in segmentation
tasks. This raises a key question: Can ControlNet additionally be guided to
generate the most informative synthetic samples for a specific task? Inspired
by active learning, where the most informative real-world samples are selected
based on sample difficulty or model uncertainty, we propose the first approach
to integrate active learning-based selection metrics into the backward
diffusion process for sample generation. Specifically, we explore uncertainty,
query by committee, and expected model change, which are commonly used in
active learning, and demonstrate their application for guiding the sample
generation process through gradient approximation. Our method is training-free,
modifying only the backward diffusion process, allowing it to be used on any
pretrained ControlNet. Using this process, we show that segmentation models
trained with guided synthetic data outperform those trained on non-guided
synthetic data. Our work underscores the need for advanced control mechanisms
for diffusion-based models, which are not only aligned with image content but
additionally downstream task performance, highlighting the true potential of
synthetic data generation.
| no_new_dataset | 0.953057 |
2503.09223 | Tian Tang | Tian Tang, Zhixing Tian, Zhenyu Zhu, Chenyang Wang, Haiqing Hu, Guoyu
Tang, Lin Liu, Sulong Xu | LREF: A Novel LLM-based Relevance Framework for E-commerce | null | null | 10.1145/3701716.3715246 | null | cs.IR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Query and product relevance prediction is a critical component for ensuring a
smooth user experience in e-commerce search. Traditional studies mainly focus
on BERT-based models to assess the semantic relevance between queries and
products. However, the discriminative paradigm and limited knowledge capacity
of these approaches restrict their ability to comprehend the relevance between
queries and products fully. With the rapid advancement of Large Language Models
(LLMs), recent research has begun to explore their application to industrial
search systems, as LLMs provide extensive world knowledge and flexible
optimization for reasoning processes. Nonetheless, directly leveraging LLMs for
relevance prediction tasks introduces new challenges, including a high demand
for data quality, the necessity for meticulous optimization of reasoning
processes, and an optimistic bias that can result in over-recall. To overcome
the above problems, this paper proposes a novel framework called the LLM-based
RElevance Framework (LREF) aimed at enhancing e-commerce search relevance. The
framework comprises three main stages: supervised fine-tuning (SFT) with Data
Selection, Multiple Chain of Thought (Multi-CoT) tuning, and Direct Preference
Optimization (DPO) for de-biasing. We evaluate the performance of the framework
through a series of offline experiments on large-scale real-world datasets, as
well as online A/B testing. The results indicate significant improvements in
both offline and online metrics. Ultimately, the model was deployed in a
well-known e-commerce application, yielding substantial commercial benefits.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:10:30 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Tang",
"Tian",
""
],
[
"Tian",
"Zhixing",
""
],
[
"Zhu",
"Zhenyu",
""
],
[
"Wang",
"Chenyang",
""
],
[
"Hu",
"Haiqing",
""
],
[
"Tang",
"Guoyu",
""
],
[
"Liu",
"Lin",
""
],
[
"Xu",
"Sulong",
""
]
]
| TITLE: LREF: A Novel LLM-based Relevance Framework for E-commerce
ABSTRACT: Query and product relevance prediction is a critical component for ensuring a
smooth user experience in e-commerce search. Traditional studies mainly focus
on BERT-based models to assess the semantic relevance between queries and
products. However, the discriminative paradigm and limited knowledge capacity
of these approaches restrict their ability to comprehend the relevance between
queries and products fully. With the rapid advancement of Large Language Models
(LLMs), recent research has begun to explore their application to industrial
search systems, as LLMs provide extensive world knowledge and flexible
optimization for reasoning processes. Nonetheless, directly leveraging LLMs for
relevance prediction tasks introduces new challenges, including a high demand
for data quality, the necessity for meticulous optimization of reasoning
processes, and an optimistic bias that can result in over-recall. To overcome
the above problems, this paper proposes a novel framework called the LLM-based
RElevance Framework (LREF) aimed at enhancing e-commerce search relevance. The
framework comprises three main stages: supervised fine-tuning (SFT) with Data
Selection, Multiple Chain of Thought (Multi-CoT) tuning, and Direct Preference
Optimization (DPO) for de-biasing. We evaluate the performance of the framework
through a series of offline experiments on large-scale real-world datasets, as
well as online A/B testing. The results indicate significant improvements in
both offline and online metrics. Ultimately, the model was deployed in a
well-known e-commerce application, yielding substantial commercial benefits.
| no_new_dataset | 0.946695 |
2503.09225 | Xuxiang Sun | Xuxiang Sun, Xianglin Shan, Yilang Liu, Weiwei Zhang | Towards a Generalized SA Model: Symbolic Regression-Based Correction for
Separated Flows | null | null | null | null | physics.flu-dyn physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study focuses on the numerical simulation of high Reynolds number
separated flows and proposes a data-driven approach to improve the predictive
capability of the SA turbulence model. First, data assimilation was performed
on two typical airfoils with high angle-of-attack separated flows to obtain a
high-fidelity flow field dataset. Based on this dataset, a white-box model was
developed using symbolic regression to modify the production term of the SA
model. To validate the effectiveness of the modified model, multiple
representative airfoils and wings, such as the SC1095 airfoil, DU91-W2-250
airfoil, and ONERA-M6 wing, were selected as test cases. A wide range of flow
conditions was considered, including subsonic to transonic regimes, Reynolds
numbers ranging from hundreds of thousands to tens of millions, and angles of
attack varying from small to large. The results indicate that the modified
model significantly improves the prediction accuracy of separated flows while
maintaining the predictive capability for attached flows. It notably enhances
the reproduction of separated vortex structures and flow separation locations,
reducing the mean relative error in lift prediction at stall angles by 69.2%
and improving computational accuracy by more than three times. Furthermore,
validation using a zero-pressure-gradient flat plate case confirms the modified
model's ability to accurately predict the turbulent boundary layer velocity
profile and skin friction coefficient distribution. The findings of this study
provide new insights and methodologies for the numerical simulation of high
Reynolds number separated flows, contributing to more accurate modeling of
complex flow phenomena in engineering applications.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:16:05 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Sun",
"Xuxiang",
""
],
[
"Shan",
"Xianglin",
""
],
[
"Liu",
"Yilang",
""
],
[
"Zhang",
"Weiwei",
""
]
]
| TITLE: Towards a Generalized SA Model: Symbolic Regression-Based Correction for
Separated Flows
ABSTRACT: This study focuses on the numerical simulation of high Reynolds number
separated flows and proposes a data-driven approach to improve the predictive
capability of the SA turbulence model. First, data assimilation was performed
on two typical airfoils with high angle-of-attack separated flows to obtain a
high-fidelity flow field dataset. Based on this dataset, a white-box model was
developed using symbolic regression to modify the production term of the SA
model. To validate the effectiveness of the modified model, multiple
representative airfoils and wings, such as the SC1095 airfoil, DU91-W2-250
airfoil, and ONERA-M6 wing, were selected as test cases. A wide range of flow
conditions was considered, including subsonic to transonic regimes, Reynolds
numbers ranging from hundreds of thousands to tens of millions, and angles of
attack varying from small to large. The results indicate that the modified
model significantly improves the prediction accuracy of separated flows while
maintaining the predictive capability for attached flows. It notably enhances
the reproduction of separated vortex structures and flow separation locations,
reducing the mean relative error in lift prediction at stall angles by 69.2%
and improving computational accuracy by more than three times. Furthermore,
validation using a zero-pressure-gradient flat plate case confirms the modified
model's ability to accurately predict the turbulent boundary layer velocity
profile and skin friction coefficient distribution. The findings of this study
provide new insights and methodologies for the numerical simulation of high
Reynolds number separated flows, contributing to more accurate modeling of
complex flow phenomena in engineering applications.
| no_new_dataset | 0.919498 |
2503.09226 | Omer Noy Klein | Omer Noy Klein, Alihan H\"uy\"uk, Ron Shamir, Uri Shalit, Mihaela van
der Schaar | Towards Regulatory-Confirmed Adaptive Clinical Trials: Machine Learning
Opportunities and Solutions | AISTATS 2025 | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Randomized Controlled Trials (RCTs) are the gold standard for evaluating the
effect of new medical treatments. Treatments must pass stringent regulatory
conditions in order to be approved for widespread use, yet even after the
regulatory barriers are crossed, real-world challenges might arise: Who should
get the treatment? What is its true clinical utility? Are there discrepancies
in the treatment effectiveness across diverse and under-served populations? We
introduce two new objectives for future clinical trials that integrate
regulatory constraints and treatment policy value for both the entire
population and under-served populations, thus answering some of the questions
above in advance. Designed to meet these objectives, we formulate Randomize
First Augment Next (RFAN), a new framework for designing Phase III clinical
trials. Our framework consists of a standard randomized component followed by
an adaptive one, jointly meant to efficiently and safely acquire and assign
patients into treatment arms during the trial. Then, we propose strategies for
implementing RFAN based on causal, deep Bayesian active learning. Finally, we
empirically evaluate the performance of our framework using synthetic and
real-world semi-synthetic datasets.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:17:54 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Klein",
"Omer Noy",
""
],
[
"Hüyük",
"Alihan",
""
],
[
"Shamir",
"Ron",
""
],
[
"Shalit",
"Uri",
""
],
[
"van der Schaar",
"Mihaela",
""
]
]
| TITLE: Towards Regulatory-Confirmed Adaptive Clinical Trials: Machine Learning
Opportunities and Solutions
ABSTRACT: Randomized Controlled Trials (RCTs) are the gold standard for evaluating the
effect of new medical treatments. Treatments must pass stringent regulatory
conditions in order to be approved for widespread use, yet even after the
regulatory barriers are crossed, real-world challenges might arise: Who should
get the treatment? What is its true clinical utility? Are there discrepancies
in the treatment effectiveness across diverse and under-served populations? We
introduce two new objectives for future clinical trials that integrate
regulatory constraints and treatment policy value for both the entire
population and under-served populations, thus answering some of the questions
above in advance. Designed to meet these objectives, we formulate Randomize
First Augment Next (RFAN), a new framework for designing Phase III clinical
trials. Our framework consists of a standard randomized component followed by
an adaptive one, jointly meant to efficiently and safely acquire and assign
patients into treatment arms during the trial. Then, we propose strategies for
implementing RFAN based on causal, deep Bayesian active learning. Finally, we
empirically evaluate the performance of our framework using synthetic and
real-world semi-synthetic datasets.
| no_new_dataset | 0.943452 |
2503.09239 | Umar Salman | Di Zhao, Umar Salman and Zongjie Wang | Risk Assessment of Distribution Networks Considering Climate Change and
Vegetation Management Impacts | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | This paper presents a comprehensive risk assessment model for power
distribution networks with a focus on the influence of climate conditions and
vegetation management on outage risks. Using a dataset comprising outage
records, meteorological indicators, and vegetation metrics, this paper develops
a logistic regression model that outperformed several alternatives, effectively
identifying risk factors in highly imbalanced data. Key features impacting
outages include wind speed, vegetation density, quantified as the enhanced
vegetation index (EVI), and snow type, with wet snow and autumn conditions
exhibiting the strongest positive effects. The analysis also shows complex
interactions, such as the combined effect of wind speed and EVI, suggesting
that vegetation density can moderate the impact of high winds on outages.
Simulation case studies, based on a test dataset of 618 samples, demonstrated
that the model achieved an 80\% match rate with real-world data within an error
tolerance of \(\pm 0.05\), showcasing the effectiveness and robustness of the
proposed model while highlighting its potential to inform preventive strategies
for mitigating outage risks in power distribution networks under high-risk
environmental conditions. Future work will integrate vegetation height data
from Lidar and explore alternative modeling approaches to capture potential
non-linear relationships.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:34:31 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zhao",
"Di",
""
],
[
"Salman",
"Umar",
""
],
[
"Wang",
"Zongjie",
""
]
]
| TITLE: Risk Assessment of Distribution Networks Considering Climate Change and
Vegetation Management Impacts
ABSTRACT: This paper presents a comprehensive risk assessment model for power
distribution networks with a focus on the influence of climate conditions and
vegetation management on outage risks. Using a dataset comprising outage
records, meteorological indicators, and vegetation metrics, this paper develops
a logistic regression model that outperformed several alternatives, effectively
identifying risk factors in highly imbalanced data. Key features impacting
outages include wind speed, vegetation density, quantified as the enhanced
vegetation index (EVI), and snow type, with wet snow and autumn conditions
exhibiting the strongest positive effects. The analysis also shows complex
interactions, such as the combined effect of wind speed and EVI, suggesting
that vegetation density can moderate the impact of high winds on outages.
Simulation case studies, based on a test dataset of 618 samples, demonstrated
that the model achieved an 80\% match rate with real-world data within an error
tolerance of \(\pm 0.05\), showcasing the effectiveness and robustness of the
proposed model while highlighting its potential to inform preventive strategies
for mitigating outage risks in power distribution networks under high-risk
environmental conditions. Future work will integrate vegetation height data
from Lidar and explore alternative modeling approaches to capture potential
non-linear relationships.
| no_new_dataset | 0.924279 |
2503.09251 | Yigang Chen | Yigang Chen, Xiang Ji, Ziyue Zhang, Yuming Zhou, Yang-Chi-Dung Lin,
Hsi-Yuan Huang, Tao Zhang, Yi Lai, Ke Chen, Chang Su, Xingqiao Lin, Zihao
Zhu, Yanggyi Zhang, Kangping Wei, Jiehui Fu, Yixian Huang, Shidong Cui,
Shih-Chung Yen, Ariel Warshel, and Hsien-Da Huang | SCOPE-DTI: Semi-Inductive Dataset Construction and Framework
Optimization for Practical Usability Enhancement in Deep Learning-Based Drug
Target Interaction Prediction | null | null | null | null | cs.LG cs.AI q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Deep learning-based drug-target interaction (DTI) prediction methods have
demonstrated strong performance; however, real-world applicability remains
constrained by limited data diversity and modeling complexity. To address these
challenges, we propose SCOPE-DTI, a unified framework combining a large-scale,
balanced semi-inductive human DTI dataset with advanced deep learning modeling.
Constructed from 13 public repositories, the SCOPE dataset expands data volume
by up to 100-fold compared to common benchmarks such as the Human dataset. The
SCOPE model integrates three-dimensional protein and compound representations,
graph neural networks, and bilinear attention mechanisms to effectively capture
cross domain interaction patterns, significantly outperforming state-of-the-art
methods across various DTI prediction tasks. Additionally, SCOPE-DTI provides a
user-friendly interface and database. We further validate its effectiveness by
experimentally identifying anticancer targets of Ginsenoside Rh1. By offering
comprehensive data, advanced modeling, and accessible tools, SCOPE-DTI
accelerates drug discovery research.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:46:25 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Chen",
"Yigang",
""
],
[
"Ji",
"Xiang",
""
],
[
"Zhang",
"Ziyue",
""
],
[
"Zhou",
"Yuming",
""
],
[
"Lin",
"Yang-Chi-Dung",
""
],
[
"Huang",
"Hsi-Yuan",
""
],
[
"Zhang",
"Tao",
""
],
[
"Lai",
"Yi",
""
],
[
"Chen",
"Ke",
""
],
[
"Su",
"Chang",
""
],
[
"Lin",
"Xingqiao",
""
],
[
"Zhu",
"Zihao",
""
],
[
"Zhang",
"Yanggyi",
""
],
[
"Wei",
"Kangping",
""
],
[
"Fu",
"Jiehui",
""
],
[
"Huang",
"Yixian",
""
],
[
"Cui",
"Shidong",
""
],
[
"Yen",
"Shih-Chung",
""
],
[
"Warshel",
"Ariel",
""
],
[
"Huang",
"Hsien-Da",
""
]
]
| TITLE: SCOPE-DTI: Semi-Inductive Dataset Construction and Framework
Optimization for Practical Usability Enhancement in Deep Learning-Based Drug
Target Interaction Prediction
ABSTRACT: Deep learning-based drug-target interaction (DTI) prediction methods have
demonstrated strong performance; however, real-world applicability remains
constrained by limited data diversity and modeling complexity. To address these
challenges, we propose SCOPE-DTI, a unified framework combining a large-scale,
balanced semi-inductive human DTI dataset with advanced deep learning modeling.
Constructed from 13 public repositories, the SCOPE dataset expands data volume
by up to 100-fold compared to common benchmarks such as the Human dataset. The
SCOPE model integrates three-dimensional protein and compound representations,
graph neural networks, and bilinear attention mechanisms to effectively capture
cross domain interaction patterns, significantly outperforming state-of-the-art
methods across various DTI prediction tasks. Additionally, SCOPE-DTI provides a
user-friendly interface and database. We further validate its effectiveness by
experimentally identifying anticancer targets of Ginsenoside Rh1. By offering
comprehensive data, advanced modeling, and accessible tools, SCOPE-DTI
accelerates drug discovery research.
| no_new_dataset | 0.930899 |
2503.09260 | Wei He | Wei He and Shangzhi Zhang and Chun-Guang Li and Xianbiao Qi and Rong
Xiao and Jun Guo | Neural Normalized Cut: A Differential and Generalizable Approach for
Spectral Clustering | 5 figures, 8 tables, accepted by Pattern Recognition (2025-03-11) | null | 10.1016/j.patcog.2025.111545 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Spectral clustering, as a popular tool for data clustering, requires an
eigen-decomposition step on a given affinity to obtain the spectral embedding.
Nevertheless, such a step suffers from the lack of generalizability and
scalability. Moreover, the obtained spectral embeddings can hardly provide a
good approximation to the ground-truth partition and thus a k-means step is
adopted to quantize the embedding. In this paper, we propose a simple yet
effective scalable and generalizable approach, called Neural Normalized Cut
(NeuNcut), to learn the clustering membership for spectral clustering directly.
In NeuNcut, we properly reparameterize the unknown cluster membership via a
neural network, and train the neural network via stochastic gradient descent
with a properly relaxed normalized cut loss. As a result, our NeuNcut enjoys a
desired generalization ability to directly infer clustering membership for
out-of-sample unseen data and hence brings us an efficient way to handle
clustering task with ultra large-scale data. We conduct extensive experiments
on both synthetic data and benchmark datasets and experimental results validate
the effectiveness and the superiority of our approach. Our code is available
at: https://github.com/hewei98/NeuNcut.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:00:16 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"He",
"Wei",
""
],
[
"Zhang",
"Shangzhi",
""
],
[
"Li",
"Chun-Guang",
""
],
[
"Qi",
"Xianbiao",
""
],
[
"Xiao",
"Rong",
""
],
[
"Guo",
"Jun",
""
]
]
| TITLE: Neural Normalized Cut: A Differential and Generalizable Approach for
Spectral Clustering
ABSTRACT: Spectral clustering, as a popular tool for data clustering, requires an
eigen-decomposition step on a given affinity to obtain the spectral embedding.
Nevertheless, such a step suffers from the lack of generalizability and
scalability. Moreover, the obtained spectral embeddings can hardly provide a
good approximation to the ground-truth partition and thus a k-means step is
adopted to quantize the embedding. In this paper, we propose a simple yet
effective scalable and generalizable approach, called Neural Normalized Cut
(NeuNcut), to learn the clustering membership for spectral clustering directly.
In NeuNcut, we properly reparameterize the unknown cluster membership via a
neural network, and train the neural network via stochastic gradient descent
with a properly relaxed normalized cut loss. As a result, our NeuNcut enjoys a
desired generalization ability to directly infer clustering membership for
out-of-sample unseen data and hence brings us an efficient way to handle
clustering task with ultra large-scale data. We conduct extensive experiments
on both synthetic data and benchmark datasets and experimental results validate
the effectiveness and the superiority of our approach. Our code is available
at: https://github.com/hewei98/NeuNcut.
| no_new_dataset | 0.94887 |
2503.09269 | Renato Portugal | Leandro C. Souza, Renato Portugal | Single-Qudit Quantum Neural Networks for Multiclass Classification | 24 pages, 3 figures, 6 tables | null | null | null | quant-ph cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper proposes a single-qudit quantum neural network for multiclass
classification, by using the enhanced representational capacity of
high-dimensional qudit states. Our design employs an $d$-dimensional unitary
operator, where $d$ corresponds to the number of classes, constructed using the
Cayley transform of a skew-symmetric matrix, to efficiently encode and process
class information. This architecture enables a direct mapping between class
labels and quantum measurement outcomes, reducing circuit depth and
computational overhead. To optimize network parameters, we introduce a hybrid
training approach that combines an extended activation function -- derived from
a truncated multivariable Taylor series expansion -- with support vector
machine optimization for weight determination. We evaluate our model on the
MNIST and EMNIST datasets, demonstrating competitive accuracy while maintaining
a compact single-qudit quantum circuit. Our findings highlight the potential of
qudit-based QNNs as scalable alternatives to classical deep learning models,
particularly for multiclass classification. However, practical implementation
remains constrained by current quantum hardware limitations. This research
advances quantum machine learning by demonstrating the feasibility of
higher-dimensional quantum systems for efficient learning tasks.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:12:05 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Souza",
"Leandro C.",
""
],
[
"Portugal",
"Renato",
""
]
]
| TITLE: Single-Qudit Quantum Neural Networks for Multiclass Classification
ABSTRACT: This paper proposes a single-qudit quantum neural network for multiclass
classification, by using the enhanced representational capacity of
high-dimensional qudit states. Our design employs an $d$-dimensional unitary
operator, where $d$ corresponds to the number of classes, constructed using the
Cayley transform of a skew-symmetric matrix, to efficiently encode and process
class information. This architecture enables a direct mapping between class
labels and quantum measurement outcomes, reducing circuit depth and
computational overhead. To optimize network parameters, we introduce a hybrid
training approach that combines an extended activation function -- derived from
a truncated multivariable Taylor series expansion -- with support vector
machine optimization for weight determination. We evaluate our model on the
MNIST and EMNIST datasets, demonstrating competitive accuracy while maintaining
a compact single-qudit quantum circuit. Our findings highlight the potential of
qudit-based QNNs as scalable alternatives to classical deep learning models,
particularly for multiclass classification. However, practical implementation
remains constrained by current quantum hardware limitations. This research
advances quantum machine learning by demonstrating the feasibility of
higher-dimensional quantum systems for efficient learning tasks.
| no_new_dataset | 0.945751 |
2503.09276 | Linzhao Jia | Linzhao Jia, Changyong Qi, Yuang Wei, Han Sun, Xiaozhe Yang | Fine-Tuning Large Language Models for Educational Support: Leveraging
Gagne's Nine Events of Instruction for Lesson Planning | null | null | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Effective lesson planning is crucial in education process, serving as the
cornerstone for high-quality teaching and the cultivation of a conducive
learning atmosphere. This study investigates how large language models (LLMs)
can enhance teacher preparation by incorporating them with Gagne's Nine Events
of Instruction, especially in the field of mathematics education in compulsory
education. It investigates two distinct methodologies: the development of Chain
of Thought (CoT) prompts to direct LLMs in generating content that aligns with
instructional events, and the application of fine-tuning approaches like
Low-Rank Adaptation (LoRA) to enhance model performance. This research starts
with creating a comprehensive dataset based on math curriculum standards and
Gagne's instructional events. The first method involves crafting CoT-optimized
prompts to generate detailed, logically coherent responses from LLMs, improving
their ability to create educationally relevant content. The second method uses
specialized datasets to fine-tune open-source models, enhancing their
educational content generation and analysis capabilities. This study
contributes to the evolving dialogue on the integration of AI in education,
illustrating innovative strategies for leveraging LLMs to bolster teaching and
learning processes.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:22:13 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Jia",
"Linzhao",
""
],
[
"Qi",
"Changyong",
""
],
[
"Wei",
"Yuang",
""
],
[
"Sun",
"Han",
""
],
[
"Yang",
"Xiaozhe",
""
]
]
| TITLE: Fine-Tuning Large Language Models for Educational Support: Leveraging
Gagne's Nine Events of Instruction for Lesson Planning
ABSTRACT: Effective lesson planning is crucial in education process, serving as the
cornerstone for high-quality teaching and the cultivation of a conducive
learning atmosphere. This study investigates how large language models (LLMs)
can enhance teacher preparation by incorporating them with Gagne's Nine Events
of Instruction, especially in the field of mathematics education in compulsory
education. It investigates two distinct methodologies: the development of Chain
of Thought (CoT) prompts to direct LLMs in generating content that aligns with
instructional events, and the application of fine-tuning approaches like
Low-Rank Adaptation (LoRA) to enhance model performance. This research starts
with creating a comprehensive dataset based on math curriculum standards and
Gagne's instructional events. The first method involves crafting CoT-optimized
prompts to generate detailed, logically coherent responses from LLMs, improving
their ability to create educationally relevant content. The second method uses
specialized datasets to fine-tune open-source models, enhancing their
educational content generation and analysis capabilities. This study
contributes to the evolving dialogue on the integration of AI in education,
illustrating innovative strategies for leveraging LLMs to bolster teaching and
learning processes.
| new_dataset | 0.956472 |
2503.09277 | Haoxuan Wang | Haoxuan Wang, Jinlong Peng, Qingdong He, Hao Yang, Ying Jin, Jiafu Wu,
Xiaobin Hu, Yanjie Pan, Zhenye Gan, Mingmin Chi, Bo Peng, Yabiao Wang | UniCombine: Unified Multi-Conditional Combination with Diffusion
Transformer | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid development of diffusion models in image generation, the
demand for more powerful and flexible controllable frameworks is increasing.
Although existing methods can guide generation beyond text prompts, the
challenge of effectively combining multiple conditional inputs while
maintaining consistency with all of them remains unsolved. To address this, we
introduce UniCombine, a DiT-based multi-conditional controllable generative
framework capable of handling any combination of conditions, including but not
limited to text prompts, spatial maps, and subject images. Specifically, we
introduce a novel Conditional MMDiT Attention mechanism and incorporate a
trainable LoRA module to build both the training-free and training-based
versions. Additionally, we propose a new pipeline to construct
SubjectSpatial200K, the first dataset designed for multi-conditional generative
tasks covering both the subject-driven and spatially-aligned conditions.
Extensive experimental results on multi-conditional generation demonstrate the
outstanding universality and powerful capability of our approach with
state-of-the-art performance.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:22:47 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Wang",
"Haoxuan",
""
],
[
"Peng",
"Jinlong",
""
],
[
"He",
"Qingdong",
""
],
[
"Yang",
"Hao",
""
],
[
"Jin",
"Ying",
""
],
[
"Wu",
"Jiafu",
""
],
[
"Hu",
"Xiaobin",
""
],
[
"Pan",
"Yanjie",
""
],
[
"Gan",
"Zhenye",
""
],
[
"Chi",
"Mingmin",
""
],
[
"Peng",
"Bo",
""
],
[
"Wang",
"Yabiao",
""
]
]
| TITLE: UniCombine: Unified Multi-Conditional Combination with Diffusion
Transformer
ABSTRACT: With the rapid development of diffusion models in image generation, the
demand for more powerful and flexible controllable frameworks is increasing.
Although existing methods can guide generation beyond text prompts, the
challenge of effectively combining multiple conditional inputs while
maintaining consistency with all of them remains unsolved. To address this, we
introduce UniCombine, a DiT-based multi-conditional controllable generative
framework capable of handling any combination of conditions, including but not
limited to text prompts, spatial maps, and subject images. Specifically, we
introduce a novel Conditional MMDiT Attention mechanism and incorporate a
trainable LoRA module to build both the training-free and training-based
versions. Additionally, we propose a new pipeline to construct
SubjectSpatial200K, the first dataset designed for multi-conditional generative
tasks covering both the subject-driven and spatially-aligned conditions.
Extensive experimental results on multi-conditional generation demonstrate the
outstanding universality and powerful capability of our approach with
state-of-the-art performance.
| new_dataset | 0.963092 |
2503.09279 | Luozheng Qin | Luozheng Qin, Zhiyu Tan, Mengping Yang, Xiaomeng Yang, Hao Li | Cockatiel: Ensembling Synthetic and Human Preferenced Training for
Detailed Video Caption | For more details, please refer to our project page:
https://sais-fuxi.github.io/projects/cockatiel/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Video Detailed Captioning (VDC) is a crucial task for vision-language
bridging, enabling fine-grained descriptions of complex video content. In this
paper, we first comprehensively benchmark current state-of-the-art approaches
and systematically identified two critical limitations: biased capability
towards specific captioning aspect and misalignment with human preferences. To
address these deficiencies, we propose Cockatiel, a novel three-stage training
pipeline that ensembles synthetic and human-aligned training for improving VDC
performance. In the first stage, we derive a scorer from a meticulously
annotated dataset to select synthetic captions high-performing on certain
fine-grained video-caption alignment and human-preferred while disregarding
others. Then, we train Cockatiel-13B, using this curated dataset to infuse it
with assembled model strengths and human preferences. Finally, we further
distill Cockatiel-8B from Cockatiel-13B for the ease of usage. Extensive
quantitative and qualitative experiments reflect the effectiveness of our
method, as we not only set new state-of-the-art performance on VDCSCORE in a
dimension-balanced way but also surpass leading alternatives on human
preference by a large margin as depicted by the human evaluation results.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:25:04 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Qin",
"Luozheng",
""
],
[
"Tan",
"Zhiyu",
""
],
[
"Yang",
"Mengping",
""
],
[
"Yang",
"Xiaomeng",
""
],
[
"Li",
"Hao",
""
]
]
| TITLE: Cockatiel: Ensembling Synthetic and Human Preferenced Training for
Detailed Video Caption
ABSTRACT: Video Detailed Captioning (VDC) is a crucial task for vision-language
bridging, enabling fine-grained descriptions of complex video content. In this
paper, we first comprehensively benchmark current state-of-the-art approaches
and systematically identified two critical limitations: biased capability
towards specific captioning aspect and misalignment with human preferences. To
address these deficiencies, we propose Cockatiel, a novel three-stage training
pipeline that ensembles synthetic and human-aligned training for improving VDC
performance. In the first stage, we derive a scorer from a meticulously
annotated dataset to select synthetic captions high-performing on certain
fine-grained video-caption alignment and human-preferred while disregarding
others. Then, we train Cockatiel-13B, using this curated dataset to infuse it
with assembled model strengths and human preferences. Finally, we further
distill Cockatiel-8B from Cockatiel-13B for the ease of usage. Extensive
quantitative and qualitative experiments reflect the effectiveness of our
method, as we not only set new state-of-the-art performance on VDCSCORE in a
dimension-balanced way but also surpass leading alternatives on human
preference by a large margin as depicted by the human evaluation results.
| new_dataset | 0.606994 |
2503.09283 | Xiangbin Wei | Xiangbin Wei | Noise2Score3D: Tweedie's Approach for Unsupervised Point Cloud Denoising | arXiv admin note: substantial text overlap with arXiv:2502.16826 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Building on recent advances in Bayesian statistics and image denoising, we
propose Noise2Score3D, a fully unsupervised framework for point cloud
denoising. Noise2Score3D learns the score function of the underlying point
cloud distribution directly from noisy data, eliminating the need for clean
data during training. Using Tweedie's formula, our method performs denoising in
a single step, avoiding the iterative processes used in existing unsupervised
methods, thus improving both accuracy and efficiency. Additionally, we
introduce Total Variation for Point Clouds as a denoising quality metric, which
allows for the estimation of unknown noise parameters. Experimental results
demonstrate that Noise2Score3D achieves state-of-the-art performance on
standard benchmarks among unsupervised learning methods in Chamfer distance and
point-to-mesh metrics. Noise2Score3D also demonstrates strong generalization
ability beyond training datasets. Our method, by addressing the generalization
issue and challenge of the absence of clean data in learning-based methods,
paves the way for learning-based point cloud denoising methods in real-world
applications.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:28:04 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Wei",
"Xiangbin",
""
]
]
| TITLE: Noise2Score3D: Tweedie's Approach for Unsupervised Point Cloud Denoising
ABSTRACT: Building on recent advances in Bayesian statistics and image denoising, we
propose Noise2Score3D, a fully unsupervised framework for point cloud
denoising. Noise2Score3D learns the score function of the underlying point
cloud distribution directly from noisy data, eliminating the need for clean
data during training. Using Tweedie's formula, our method performs denoising in
a single step, avoiding the iterative processes used in existing unsupervised
methods, thus improving both accuracy and efficiency. Additionally, we
introduce Total Variation for Point Clouds as a denoising quality metric, which
allows for the estimation of unknown noise parameters. Experimental results
demonstrate that Noise2Score3D achieves state-of-the-art performance on
standard benchmarks among unsupervised learning methods in Chamfer distance and
point-to-mesh metrics. Noise2Score3D also demonstrates strong generalization
ability beyond training datasets. Our method, by addressing the generalization
issue and challenge of the absence of clean data in learning-based methods,
paves the way for learning-based point cloud denoising methods in real-world
applications.
| no_new_dataset | 0.94801 |
2503.09291 | Xinjian Luo | Xinjian Luo, Ting Yu, Xiaokui Xiao | Prompt Inference Attack on Distributed Large Language Model Inference
Frameworks | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The inference process of modern large language models (LLMs) demands
prohibitive computational resources, rendering them infeasible for deployment
on consumer-grade devices. To address this limitation, recent studies propose
distributed LLM inference frameworks, which employ split learning principles to
enable collaborative LLM inference on resource-constrained hardware. However,
distributing LLM layers across participants requires the transmission of
intermediate outputs, which may introduce privacy risks to the original input
prompts - a critical issue that has yet to be thoroughly explored in the
literature.
In this paper, we rigorously examine the privacy vulnerabilities of
distributed LLM inference frameworks by designing and evaluating three prompt
inference attacks aimed at reconstructing input prompts from intermediate LLM
outputs. These attacks are developed under various query and data constraints
to reflect diverse real-world LLM service scenarios. Specifically, the first
attack assumes an unlimited query budget and access to an auxiliary dataset
sharing the same distribution as the target prompts. The second attack also
leverages unlimited queries but uses an auxiliary dataset with a distribution
differing from the target prompts. The third attack operates under the most
restrictive scenario, with limited query budgets and no auxiliary dataset
available. We evaluate these attacks on a range of LLMs, including
state-of-the-art models such as Llama-3.2 and Phi-3.5, as well as widely-used
models like GPT-2 and BERT for comparative analysis. Our experiments show that
the first two attacks achieve reconstruction accuracies exceeding 90%, while
the third achieves accuracies typically above 50%, even under stringent
constraints. These findings highlight privacy risks in distributed LLM
inference frameworks, issuing a strong alert on their deployment in real-world
applications.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:36:29 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Luo",
"Xinjian",
""
],
[
"Yu",
"Ting",
""
],
[
"Xiao",
"Xiaokui",
""
]
]
| TITLE: Prompt Inference Attack on Distributed Large Language Model Inference
Frameworks
ABSTRACT: The inference process of modern large language models (LLMs) demands
prohibitive computational resources, rendering them infeasible for deployment
on consumer-grade devices. To address this limitation, recent studies propose
distributed LLM inference frameworks, which employ split learning principles to
enable collaborative LLM inference on resource-constrained hardware. However,
distributing LLM layers across participants requires the transmission of
intermediate outputs, which may introduce privacy risks to the original input
prompts - a critical issue that has yet to be thoroughly explored in the
literature.
In this paper, we rigorously examine the privacy vulnerabilities of
distributed LLM inference frameworks by designing and evaluating three prompt
inference attacks aimed at reconstructing input prompts from intermediate LLM
outputs. These attacks are developed under various query and data constraints
to reflect diverse real-world LLM service scenarios. Specifically, the first
attack assumes an unlimited query budget and access to an auxiliary dataset
sharing the same distribution as the target prompts. The second attack also
leverages unlimited queries but uses an auxiliary dataset with a distribution
differing from the target prompts. The third attack operates under the most
restrictive scenario, with limited query budgets and no auxiliary dataset
available. We evaluate these attacks on a range of LLMs, including
state-of-the-art models such as Llama-3.2 and Phi-3.5, as well as widely-used
models like GPT-2 and BERT for comparative analysis. Our experiments show that
the first two attacks achieve reconstruction accuracies exceeding 90%, while
the third achieves accuracies typically above 50%, even under stringent
constraints. These findings highlight privacy risks in distributed LLM
inference frameworks, issuing a strong alert on their deployment in real-world
applications.
| no_new_dataset | 0.941439 |
2503.09294 | Peng Hu | Peng Hu, Chunming He, Lei Xu, Jingduo Tian, Sina Farsiu, Yulun Zhang,
Pei Liu and Xiu Li | IQPFR: An Image Quality Prior for Blind Face Restoration and Beyond | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Blind Face Restoration (BFR) addresses the challenge of reconstructing
degraded low-quality (LQ) facial images into high-quality (HQ) outputs.
Conventional approaches predominantly rely on learning feature representations
from ground-truth (GT) data; however, inherent imperfections in GT datasets
constrain restoration performance to the mean quality level of the training
data, rather than attaining maximally attainable visual quality. To overcome
this limitation, we propose a novel framework that incorporates an Image
Quality Prior (IQP) derived from No-Reference Image Quality Assessment (NR-IQA)
models to guide the restoration process toward optimal HQ reconstructions. Our
methodology synergizes this IQP with a learned codebook prior through two
critical innovations: (1) During codebook learning, we devise a dual-branch
codebook architecture that disentangles feature extraction into universal
structural components and HQ-specific attributes, ensuring comprehensive
representation of both common and high-quality facial characteristics. (2) In
the codebook lookup stage, we implement a quality-conditioned Transformer-based
framework. NR-IQA-derived quality scores act as dynamic conditioning signals to
steer restoration toward the highest feasible quality standard. This
score-conditioned paradigm enables plug-and-play enhancement of existing BFR
architectures without modifying the original structure. We also formulate a
discrete representation-based quality optimization strategy that circumvents
over-optimization artifacts prevalent in continuous latent space approaches.
Extensive experiments demonstrate that our method outperforms state-of-the-art
techniques across multiple benchmarks. Besides, our quality-conditioned
framework demonstrates consistent performance improvements when integrated with
prior BFR models. The code will be released.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:39:51 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Hu",
"Peng",
""
],
[
"He",
"Chunming",
""
],
[
"Xu",
"Lei",
""
],
[
"Tian",
"Jingduo",
""
],
[
"Farsiu",
"Sina",
""
],
[
"Zhang",
"Yulun",
""
],
[
"Liu",
"Pei",
""
],
[
"Li",
"Xiu",
""
]
]
| TITLE: IQPFR: An Image Quality Prior for Blind Face Restoration and Beyond
ABSTRACT: Blind Face Restoration (BFR) addresses the challenge of reconstructing
degraded low-quality (LQ) facial images into high-quality (HQ) outputs.
Conventional approaches predominantly rely on learning feature representations
from ground-truth (GT) data; however, inherent imperfections in GT datasets
constrain restoration performance to the mean quality level of the training
data, rather than attaining maximally attainable visual quality. To overcome
this limitation, we propose a novel framework that incorporates an Image
Quality Prior (IQP) derived from No-Reference Image Quality Assessment (NR-IQA)
models to guide the restoration process toward optimal HQ reconstructions. Our
methodology synergizes this IQP with a learned codebook prior through two
critical innovations: (1) During codebook learning, we devise a dual-branch
codebook architecture that disentangles feature extraction into universal
structural components and HQ-specific attributes, ensuring comprehensive
representation of both common and high-quality facial characteristics. (2) In
the codebook lookup stage, we implement a quality-conditioned Transformer-based
framework. NR-IQA-derived quality scores act as dynamic conditioning signals to
steer restoration toward the highest feasible quality standard. This
score-conditioned paradigm enables plug-and-play enhancement of existing BFR
architectures without modifying the original structure. We also formulate a
discrete representation-based quality optimization strategy that circumvents
over-optimization artifacts prevalent in continuous latent space approaches.
Extensive experiments demonstrate that our method outperforms state-of-the-art
techniques across multiple benchmarks. Besides, our quality-conditioned
framework demonstrates consistent performance improvements when integrated with
prior BFR models. The code will be released.
| no_new_dataset | 0.945197 |
2503.09296 | Bingzheng Jiang | Bingzheng Jiang, Jiayuan Wang, Han Ding, Lijun Zhu | MonoSLAM: Robust Monocular SLAM with Global Structure Optimization | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a robust monocular visual SLAM system that simultaneously
utilizes point, line, and vanishing point features for accurate camera pose
estimation and mapping. To address the critical challenge of achieving reliable
localization in low-texture environments, where traditional point-based systems
often fail due to insufficient visual features, we introduce a novel approach
leveraging Global Primitives structural information to improve the system's
robustness and accuracy performance. Our key innovation lies in constructing
vanishing points from line features and proposing a weighted fusion strategy to
build Global Primitives in the world coordinate system. This strategy
associates multiple frames with non-overlapping regions and formulates a
multi-frame reprojection error optimization, significantly improving tracking
accuracy in texture-scarce scenarios. Evaluations on various datasets show that
our system outperforms state-of-the-art methods in trajectory precision,
particularly in challenging environments.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:43:10 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Jiang",
"Bingzheng",
""
],
[
"Wang",
"Jiayuan",
""
],
[
"Ding",
"Han",
""
],
[
"Zhu",
"Lijun",
""
]
]
| TITLE: MonoSLAM: Robust Monocular SLAM with Global Structure Optimization
ABSTRACT: This paper presents a robust monocular visual SLAM system that simultaneously
utilizes point, line, and vanishing point features for accurate camera pose
estimation and mapping. To address the critical challenge of achieving reliable
localization in low-texture environments, where traditional point-based systems
often fail due to insufficient visual features, we introduce a novel approach
leveraging Global Primitives structural information to improve the system's
robustness and accuracy performance. Our key innovation lies in constructing
vanishing points from line features and proposing a weighted fusion strategy to
build Global Primitives in the world coordinate system. This strategy
associates multiple frames with non-overlapping regions and formulates a
multi-frame reprojection error optimization, significantly improving tracking
accuracy in texture-scarce scenarios. Evaluations on various datasets show that
our system outperforms state-of-the-art methods in trajectory precision,
particularly in challenging environments.
| no_new_dataset | 0.948537 |
2503.09302 | Augustine O. Nwajana | Halima I. Kure, Pradipta Sarkar, Ahmed B. Ndanusa, and Augustine O.
Nwajana | Detecting and Preventing Data Poisoning Attacks on AI Models | 9 pages, 8 figures | null | null | null | cs.CR eess.IV | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the critical issue of data poisoning attacks on AI
models, a growing concern in the ever-evolving landscape of artificial
intelligence and cybersecurity. As advanced technology systems become
increasingly prevalent across various sectors, the need for robust defence
mechanisms against adversarial attacks becomes paramount. The study aims to
develop and evaluate novel techniques for detecting and preventing data
poisoning attacks, focusing on both theoretical frameworks and practical
applications. Through a comprehensive literature review, experimental
validation using the CIFAR-10 and Insurance Claims datasets, and the
development of innovative algorithms, this paper seeks to enhance the
resilience of AI models against malicious data manipulation. The study explores
various methods, including anomaly detection, robust optimization strategies,
and ensemble learning, to identify and mitigate the effects of poisoned data
during model training. Experimental results indicate that data poisoning
significantly degrades model performance, reducing classification accuracy by
up to 27% in image recognition tasks (CIFAR-10) and 22% in fraud detection
models (Insurance Claims dataset). The proposed defence mechanisms, including
statistical anomaly detection and adversarial training, successfully mitigated
poisoning effects, improving model robustness and restoring accuracy levels by
an average of 15-20%. The findings further demonstrate that ensemble learning
techniques provide an additional layer of resilience, reducing false positives
and false negatives caused by adversarial data injections.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:55:01 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Kure",
"Halima I.",
""
],
[
"Sarkar",
"Pradipta",
""
],
[
"Ndanusa",
"Ahmed B.",
""
],
[
"Nwajana",
"Augustine O.",
""
]
]
| TITLE: Detecting and Preventing Data Poisoning Attacks on AI Models
ABSTRACT: This paper investigates the critical issue of data poisoning attacks on AI
models, a growing concern in the ever-evolving landscape of artificial
intelligence and cybersecurity. As advanced technology systems become
increasingly prevalent across various sectors, the need for robust defence
mechanisms against adversarial attacks becomes paramount. The study aims to
develop and evaluate novel techniques for detecting and preventing data
poisoning attacks, focusing on both theoretical frameworks and practical
applications. Through a comprehensive literature review, experimental
validation using the CIFAR-10 and Insurance Claims datasets, and the
development of innovative algorithms, this paper seeks to enhance the
resilience of AI models against malicious data manipulation. The study explores
various methods, including anomaly detection, robust optimization strategies,
and ensemble learning, to identify and mitigate the effects of poisoned data
during model training. Experimental results indicate that data poisoning
significantly degrades model performance, reducing classification accuracy by
up to 27% in image recognition tasks (CIFAR-10) and 22% in fraud detection
models (Insurance Claims dataset). The proposed defence mechanisms, including
statistical anomaly detection and adversarial training, successfully mitigated
poisoning effects, improving model robustness and restoring accuracy levels by
an average of 15-20%. The findings further demonstrate that ensemble learning
techniques provide an additional layer of resilience, reducing false positives
and false negatives caused by adversarial data injections.
| no_new_dataset | 0.946151 |
2503.09321 | Gorjan Radevski | Gorjan Radevski, Teodora Popordanoska, Matthew B. Blaschko, Tinne
Tuytelaars | DAVE: Diagnostic benchmark for Audio Visual Evaluation | First two authors contributed equally | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Audio-visual understanding is a rapidly evolving field that seeks to
integrate and interpret information from both auditory and visual modalities.
Despite recent advances in multi-modal learning, existing benchmarks often
suffer from strong visual bias -- where answers can be inferred from visual
data alone -- and provide only aggregate scores that conflate multiple sources
of error. This makes it difficult to determine whether models struggle with
visual understanding, audio interpretation, or audio-visual alignment. In this
work, we introduce DAVE (Diagnostic Audio Visual Evaluation), a novel benchmark
dataset designed to systematically evaluate audio-visual models across
controlled challenges. DAVE alleviates existing limitations by (i) ensuring
both modalities are necessary to answer correctly and (ii) decoupling
evaluation into atomic subcategories. Our detailed analysis of state-of-the-art
models reveals specific failure modes and provides targeted insights for
improvement. By offering this standardized diagnostic framework, we aim to
facilitate more robust development of audio-visual models. The dataset is
released: https://github.com/gorjanradevski/dave
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:12:46 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Radevski",
"Gorjan",
""
],
[
"Popordanoska",
"Teodora",
""
],
[
"Blaschko",
"Matthew B.",
""
],
[
"Tuytelaars",
"Tinne",
""
]
]
| TITLE: DAVE: Diagnostic benchmark for Audio Visual Evaluation
ABSTRACT: Audio-visual understanding is a rapidly evolving field that seeks to
integrate and interpret information from both auditory and visual modalities.
Despite recent advances in multi-modal learning, existing benchmarks often
suffer from strong visual bias -- where answers can be inferred from visual
data alone -- and provide only aggregate scores that conflate multiple sources
of error. This makes it difficult to determine whether models struggle with
visual understanding, audio interpretation, or audio-visual alignment. In this
work, we introduce DAVE (Diagnostic Audio Visual Evaluation), a novel benchmark
dataset designed to systematically evaluate audio-visual models across
controlled challenges. DAVE alleviates existing limitations by (i) ensuring
both modalities are necessary to answer correctly and (ii) decoupling
evaluation into atomic subcategories. Our detailed analysis of state-of-the-art
models reveals specific failure modes and provides targeted insights for
improvement. By offering this standardized diagnostic framework, we aim to
facilitate more robust development of audio-visual models. The dataset is
released: https://github.com/gorjanradevski/dave
| new_dataset | 0.957118 |
2503.09330 | Thomas De Min | Thomas De Min, Subhankar Roy, St\'ephane Lathuili\`ere, Elisa Ricci,
Massimiliano Mancini | Group-robust Machine Unlearning | Work in progress | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Machine unlearning is an emerging paradigm to remove the influence of
specific training data (i.e., the forget set) from a model while preserving its
knowledge of the rest of the data (i.e., the retain set). Previous approaches
assume the forget data to be uniformly distributed from all training
datapoints. However, if the data to unlearn is dominant in one group, we
empirically show that performance for this group degrades, leading to fairness
issues. This work tackles the overlooked problem of non-uniformly distributed
forget sets, which we call group-robust machine unlearning, by presenting a
simple, effective strategy that mitigates the performance loss in dominant
groups via sample distribution reweighting. Moreover, we present MIU (Mutual
Information-aware Machine Unlearning), the first approach for group robustness
in approximate machine unlearning. MIU minimizes the mutual information between
model features and group information, achieving unlearning while reducing
performance degradation in the dominant group of the forget set. Additionally,
MIU exploits sample distribution reweighting and mutual information calibration
with the original model to preserve group robustness. We conduct experiments on
three datasets and show that MIU outperforms standard methods, achieving
unlearning without compromising model robustness. Source code available at
https://github.com/tdemin16/group-robust_machine_unlearning.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:24:05 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"De Min",
"Thomas",
""
],
[
"Roy",
"Subhankar",
""
],
[
"Lathuilière",
"Stéphane",
""
],
[
"Ricci",
"Elisa",
""
],
[
"Mancini",
"Massimiliano",
""
]
]
| TITLE: Group-robust Machine Unlearning
ABSTRACT: Machine unlearning is an emerging paradigm to remove the influence of
specific training data (i.e., the forget set) from a model while preserving its
knowledge of the rest of the data (i.e., the retain set). Previous approaches
assume the forget data to be uniformly distributed from all training
datapoints. However, if the data to unlearn is dominant in one group, we
empirically show that performance for this group degrades, leading to fairness
issues. This work tackles the overlooked problem of non-uniformly distributed
forget sets, which we call group-robust machine unlearning, by presenting a
simple, effective strategy that mitigates the performance loss in dominant
groups via sample distribution reweighting. Moreover, we present MIU (Mutual
Information-aware Machine Unlearning), the first approach for group robustness
in approximate machine unlearning. MIU minimizes the mutual information between
model features and group information, achieving unlearning while reducing
performance degradation in the dominant group of the forget set. Additionally,
MIU exploits sample distribution reweighting and mutual information calibration
with the original model to preserve group robustness. We conduct experiments on
three datasets and show that MIU outperforms standard methods, achieving
unlearning without compromising model robustness. Source code available at
https://github.com/tdemin16/group-robust_machine_unlearning.
| no_new_dataset | 0.948775 |
2503.09332 | Dai Sun | Dai Sun and Huhao Guan and Kun Zhang and Xike Xie and S. Kevin Zhou | SDD-4DGS: Static-Dynamic Aware Decoupling in Gaussian Splatting for 4D
Scene Reconstruction | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Dynamic and static components in scenes often exhibit distinct properties,
yet most 4D reconstruction methods treat them indiscriminately, leading to
suboptimal performance in both cases. This work introduces SDD-4DGS, the first
framework for static-dynamic decoupled 4D scene reconstruction based on
Gaussian Splatting. Our approach is built upon a novel probabilistic dynamic
perception coefficient that is naturally integrated into the Gaussian
reconstruction pipeline, enabling adaptive separation of static and dynamic
components. With carefully designed implementation strategies to realize this
theoretical framework, our method effectively facilitates explicit learning of
motion patterns for dynamic elements while maintaining geometric stability for
static structures. Extensive experiments on five benchmark datasets demonstrate
that SDD-4DGS consistently outperforms state-of-the-art methods in
reconstruction fidelity, with enhanced detail restoration for static structures
and precise modeling of dynamic motions. The code will be released.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:25:58 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Sun",
"Dai",
""
],
[
"Guan",
"Huhao",
""
],
[
"Zhang",
"Kun",
""
],
[
"Xie",
"Xike",
""
],
[
"Zhou",
"S. Kevin",
""
]
]
| TITLE: SDD-4DGS: Static-Dynamic Aware Decoupling in Gaussian Splatting for 4D
Scene Reconstruction
ABSTRACT: Dynamic and static components in scenes often exhibit distinct properties,
yet most 4D reconstruction methods treat them indiscriminately, leading to
suboptimal performance in both cases. This work introduces SDD-4DGS, the first
framework for static-dynamic decoupled 4D scene reconstruction based on
Gaussian Splatting. Our approach is built upon a novel probabilistic dynamic
perception coefficient that is naturally integrated into the Gaussian
reconstruction pipeline, enabling adaptive separation of static and dynamic
components. With carefully designed implementation strategies to realize this
theoretical framework, our method effectively facilitates explicit learning of
motion patterns for dynamic elements while maintaining geometric stability for
static structures. Extensive experiments on five benchmark datasets demonstrate
that SDD-4DGS consistently outperforms state-of-the-art methods in
reconstruction fidelity, with enhanced detail restoration for static structures
and precise modeling of dynamic motions. The code will be released.
| no_new_dataset | 0.943034 |
2503.09342 | Aymen Mir | Aymen Mir, Arthur Moreau, Helisa Dhamo, Zhensong Zhang, Eduardo
P\'erez-Pellitero | GASPACHO: Gaussian Splatting for Controllable Humans and Objects | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present GASPACHO: a method for generating photorealistic controllable
renderings of human-object interactions. Given a set of multi-view RGB images
of human-object interactions, our method reconstructs animatable templates of
the human and object as separate sets of Gaussians simultaneously. Different
from existing work, which focuses on human reconstruction and ignores objects
as background, our method explicitly reconstructs both humans and objects,
thereby allowing for controllable renderings of novel human object interactions
in different poses from novel-camera viewpoints. During reconstruction, we
constrain the Gaussians that generate rendered images to be a linear function
of a set of canonical Gaussians. By simply changing the parameters of the
linear deformation functions after training, our method can generate renderings
of novel human-object interaction in novel poses from novel camera viewpoints.
We learn the 3D Gaussian properties of the canonical Gaussians on the
underlying 2D manifold of the canonical human and object templates. This in
turns requires a canonical object template with a fixed UV unwrapping. To
define such an object template, we use a feature based representation to track
the object across the multi-view sequence. We further propose an occlusion
aware photometric loss that allows for reconstructions under significant
occlusions. Several experiments on two human-object datasets - BEHAVE and
DNA-Rendering - demonstrate that our method allows for high-quality
reconstruction of human and object templates under significant occlusion and
the synthesis of controllable renderings of novel human-object interactions in
novel human poses from novel camera views.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:38:48 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Mir",
"Aymen",
""
],
[
"Moreau",
"Arthur",
""
],
[
"Dhamo",
"Helisa",
""
],
[
"Zhang",
"Zhensong",
""
],
[
"Pérez-Pellitero",
"Eduardo",
""
]
]
| TITLE: GASPACHO: Gaussian Splatting for Controllable Humans and Objects
ABSTRACT: We present GASPACHO: a method for generating photorealistic controllable
renderings of human-object interactions. Given a set of multi-view RGB images
of human-object interactions, our method reconstructs animatable templates of
the human and object as separate sets of Gaussians simultaneously. Different
from existing work, which focuses on human reconstruction and ignores objects
as background, our method explicitly reconstructs both humans and objects,
thereby allowing for controllable renderings of novel human object interactions
in different poses from novel-camera viewpoints. During reconstruction, we
constrain the Gaussians that generate rendered images to be a linear function
of a set of canonical Gaussians. By simply changing the parameters of the
linear deformation functions after training, our method can generate renderings
of novel human-object interaction in novel poses from novel camera viewpoints.
We learn the 3D Gaussian properties of the canonical Gaussians on the
underlying 2D manifold of the canonical human and object templates. This in
turns requires a canonical object template with a fixed UV unwrapping. To
define such an object template, we use a feature based representation to track
the object across the multi-view sequence. We further propose an occlusion
aware photometric loss that allows for reconstructions under significant
occlusions. Several experiments on two human-object datasets - BEHAVE and
DNA-Rendering - demonstrate that our method allows for high-quality
reconstruction of human and object templates under significant occlusion and
the synthesis of controllable renderings of novel human-object interactions in
novel human poses from novel camera views.
| no_new_dataset | 0.954732 |
2503.09344 | Lu Qi | Lehan Yang, Lu Qi, Xiangtai Li, Sheng Li, Varun Jampani, Ming-Hsuan
Yang | Unified Dense Prediction of Video Diffusion | Accepted by CVPR2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present a unified network for simultaneously generating videos and their
corresponding entity segmentation and depth maps from text prompts. We utilize
colormap to represent entity masks and depth maps, tightly integrating dense
prediction with RGB video generation. Introducing dense prediction information
improves video generation's consistency and motion smoothness without
increasing computational costs. Incorporating learnable task embeddings brings
multiple dense prediction tasks into a single model, enhancing flexibility and
further boosting performance. We further propose a large-scale dense prediction
video dataset~\datasetname, addressing the issue that existing datasets do not
concurrently contain captions, videos, segmentation, or depth maps.
Comprehensive experiments demonstrate the high efficiency of our method,
surpassing the state-of-the-art in terms of video quality, consistency, and
motion smoothness.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:41:02 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Yang",
"Lehan",
""
],
[
"Qi",
"Lu",
""
],
[
"Li",
"Xiangtai",
""
],
[
"Li",
"Sheng",
""
],
[
"Jampani",
"Varun",
""
],
[
"Yang",
"Ming-Hsuan",
""
]
]
| TITLE: Unified Dense Prediction of Video Diffusion
ABSTRACT: We present a unified network for simultaneously generating videos and their
corresponding entity segmentation and depth maps from text prompts. We utilize
colormap to represent entity masks and depth maps, tightly integrating dense
prediction with RGB video generation. Introducing dense prediction information
improves video generation's consistency and motion smoothness without
increasing computational costs. Incorporating learnable task embeddings brings
multiple dense prediction tasks into a single model, enhancing flexibility and
further boosting performance. We further propose a large-scale dense prediction
video dataset~\datasetname, addressing the issue that existing datasets do not
concurrently contain captions, videos, segmentation, or depth maps.
Comprehensive experiments demonstrate the high efficiency of our method,
surpassing the state-of-the-art in terms of video quality, consistency, and
motion smoothness.
| new_dataset | 0.945851 |
2503.09349 | Simon Geirnaert | Simon Geirnaert, Jonas Vanthornhout, Tom Francart, Alexander Bertrand | Performance Modeling for Correlation-based Neural Decoding of Auditory
Attention to Speech | null | null | null | null | eess.SP cs.SD eess.AS q-bio.NC | http://creativecommons.org/licenses/by/4.0/ | Correlation-based auditory attention decoding (AAD) algorithms exploit neural
tracking mechanisms to determine listener attention among competing speech
sources via, e.g., electroencephalography signals. The correlation coefficients
between the decoded neural responses and encoded speech stimuli of the
different speakers then serve as AAD decision variables. A critical trade-off
exists between the temporal resolution (the decision window length used to
compute these correlations) and the AAD accuracy. This trade-off is typically
characterized by evaluating AAD accuracy across multiple window lengths,
leading to the performance curve. We propose a novel method to model this
trade-off curve using labeled correlations from only a single decision window
length. Our approach models the (un)attended correlations with a normal
distribution after applying the Fisher transformation, enabling accurate AAD
accuracy prediction across different window lengths. We validate the method on
two distinct AAD implementations: a linear decoder and the non-linear VLAAI
deep neural network, evaluated on separate datasets. Results show consistently
low modeling errors of approximately 2 percent points, with 94% of true
accuracies falling within estimated 95%-confidence intervals. The proposed
method enables efficient performance curve modeling without extensive
multi-window length evaluation, facilitating practical applications in, e.g.,
performance tracking in neuro-steered hearing devices to continuously adapt the
system parameters over time.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:51:46 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Geirnaert",
"Simon",
""
],
[
"Vanthornhout",
"Jonas",
""
],
[
"Francart",
"Tom",
""
],
[
"Bertrand",
"Alexander",
""
]
]
| TITLE: Performance Modeling for Correlation-based Neural Decoding of Auditory
Attention to Speech
ABSTRACT: Correlation-based auditory attention decoding (AAD) algorithms exploit neural
tracking mechanisms to determine listener attention among competing speech
sources via, e.g., electroencephalography signals. The correlation coefficients
between the decoded neural responses and encoded speech stimuli of the
different speakers then serve as AAD decision variables. A critical trade-off
exists between the temporal resolution (the decision window length used to
compute these correlations) and the AAD accuracy. This trade-off is typically
characterized by evaluating AAD accuracy across multiple window lengths,
leading to the performance curve. We propose a novel method to model this
trade-off curve using labeled correlations from only a single decision window
length. Our approach models the (un)attended correlations with a normal
distribution after applying the Fisher transformation, enabling accurate AAD
accuracy prediction across different window lengths. We validate the method on
two distinct AAD implementations: a linear decoder and the non-linear VLAAI
deep neural network, evaluated on separate datasets. Results show consistently
low modeling errors of approximately 2 percent points, with 94% of true
accuracies falling within estimated 95%-confidence intervals. The proposed
method enables efficient performance curve modeling without extensive
multi-window length evaluation, facilitating practical applications in, e.g.,
performance tracking in neuro-steered hearing devices to continuously adapt the
system parameters over time.
| no_new_dataset | 0.948394 |
2503.09354 | Christoph Huber | Christoph Huber, Dino Knoll and Michael Guthe | Fully-Synthetic Training for Visual Quality Inspection in Automotive
Production | Accepted for publication in Procedia CIRP | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Visual Quality Inspection plays a crucial role in modern manufacturing
environments as it ensures customer safety and satisfaction. The introduction
of Computer Vision (CV) has revolutionized visual quality inspection by
improving the accuracy and efficiency of defect detection. However, traditional
CV models heavily rely on extensive datasets for training, which can be costly,
time-consuming, and error-prone. To overcome these challenges, synthetic images
have emerged as a promising alternative. They offer a cost-effective solution
with automatically generated labels. In this paper, we propose a pipeline for
generating synthetic images using domain randomization. We evaluate our
approach in three real inspection scenarios and demonstrate that an object
detection model trained solely on synthetic data can outperform models trained
on real images.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:58:30 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Huber",
"Christoph",
""
],
[
"Knoll",
"Dino",
""
],
[
"Guthe",
"Michael",
""
]
]
| TITLE: Fully-Synthetic Training for Visual Quality Inspection in Automotive
Production
ABSTRACT: Visual Quality Inspection plays a crucial role in modern manufacturing
environments as it ensures customer safety and satisfaction. The introduction
of Computer Vision (CV) has revolutionized visual quality inspection by
improving the accuracy and efficiency of defect detection. However, traditional
CV models heavily rely on extensive datasets for training, which can be costly,
time-consuming, and error-prone. To overcome these challenges, synthetic images
have emerged as a promising alternative. They offer a cost-effective solution
with automatically generated labels. In this paper, we propose a pipeline for
generating synthetic images using domain randomization. We evaluate our
approach in three real inspection scenarios and demonstrate that an object
detection model trained solely on synthetic data can outperform models trained
on real images.
| no_new_dataset | 0.953362 |
2503.09355 | Lianyuan Yu | Lianyuan Yu, Xiuzhen Guo, Ji Shi, Hongxiao Wang, and Hongwei Li | GIGP: A Global Information Interacting and Geometric Priors Focusing
Framework for Semi-supervised Medical Image Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semi-supervised learning enhances medical image segmentation by leveraging
unlabeled data, reducing reliance on extensive labeled datasets. On the one
hand, the distribution discrepancy between limited labeled data and abundant
unlabeled data can hinder model generalization. Most existing methods rely on
local similarity matching, which may introduce bias. In contrast, Mamba
effectively models global context with linear complexity, learning more
comprehensive data representations. On the other hand, medical images usually
exhibit consistent anatomical structures defined by geometric features. Most
existing methods fail to fully utilize global geometric priors, such as
volumes, moments etc. In this work, we introduce a global information
interaction and geometric priors focus framework (GIGP). Firstly, we present a
Global Information Interaction Mamba module to reduce distribution discrepancy
between labeled and unlabeled data. Secondly, we propose a Geometric Moment
Attention Mechanism to extract richer global geometric features. Finally, we
propose Global Geometric Perturbation Consistency to simulate organ dynamics
and geometric variations, enhancing the ability of the model to learn
generalized features. The superior performance on the NIH Pancreas and Left
Atrium datasets demonstrates the effectiveness of our approach.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:59:38 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Yu",
"Lianyuan",
""
],
[
"Guo",
"Xiuzhen",
""
],
[
"Shi",
"Ji",
""
],
[
"Wang",
"Hongxiao",
""
],
[
"Li",
"Hongwei",
""
]
]
| TITLE: GIGP: A Global Information Interacting and Geometric Priors Focusing
Framework for Semi-supervised Medical Image Segmentation
ABSTRACT: Semi-supervised learning enhances medical image segmentation by leveraging
unlabeled data, reducing reliance on extensive labeled datasets. On the one
hand, the distribution discrepancy between limited labeled data and abundant
unlabeled data can hinder model generalization. Most existing methods rely on
local similarity matching, which may introduce bias. In contrast, Mamba
effectively models global context with linear complexity, learning more
comprehensive data representations. On the other hand, medical images usually
exhibit consistent anatomical structures defined by geometric features. Most
existing methods fail to fully utilize global geometric priors, such as
volumes, moments etc. In this work, we introduce a global information
interaction and geometric priors focus framework (GIGP). Firstly, we present a
Global Information Interaction Mamba module to reduce distribution discrepancy
between labeled and unlabeled data. Secondly, we propose a Geometric Moment
Attention Mechanism to extract richer global geometric features. Finally, we
propose Global Geometric Perturbation Consistency to simulate organ dynamics
and geometric variations, enhancing the ability of the model to learn
generalized features. The superior performance on the NIH Pancreas and Left
Atrium datasets demonstrates the effectiveness of our approach.
| no_new_dataset | 0.951774 |
2503.09358 | Jiushen Cai | Jiushen Cai, Weihang Zhang, Hanruo Liu, Ningli Wang, and Huiqi Li | RetSTA: An LLM-Based Approach for Standardizing Clinical Fundus Image
Reports | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Standardization of clinical reports is crucial for improving the quality of
healthcare and facilitating data integration. The lack of unified standards,
including format, terminology, and style, is a great challenge in clinical
fundus diagnostic reports, which increases the difficulty for large language
models (LLMs) to understand the data. To address this, we construct a bilingual
standard terminology, containing fundus clinical terms and commonly used
descriptions in clinical diagnosis. Then, we establish two models,
RetSTA-7B-Zero and RetSTA-7B. RetSTA-7B-Zero, fine-tuned on an augmented
dataset simulating clinical scenarios, demonstrates powerful standardization
behaviors. However, it encounters a challenge of limitation to cover a wider
range of diseases. To further enhance standardization performance, we build
RetSTA-7B, which integrates a substantial amount of standardized data generated
by RetSTA-7B-Zero along with corresponding English data, covering diverse
complex clinical scenarios and achieving report-level standardization for the
first time. Experimental results demonstrate that RetSTA-7B outperforms other
compared LLMs in bilingual standardization task, which validates its superior
performance and generalizability. The checkpoints are available at
https://github.com/AB-Story/RetSTA-7B.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:00:57 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Cai",
"Jiushen",
""
],
[
"Zhang",
"Weihang",
""
],
[
"Liu",
"Hanruo",
""
],
[
"Wang",
"Ningli",
""
],
[
"Li",
"Huiqi",
""
]
]
| TITLE: RetSTA: An LLM-Based Approach for Standardizing Clinical Fundus Image
Reports
ABSTRACT: Standardization of clinical reports is crucial for improving the quality of
healthcare and facilitating data integration. The lack of unified standards,
including format, terminology, and style, is a great challenge in clinical
fundus diagnostic reports, which increases the difficulty for large language
models (LLMs) to understand the data. To address this, we construct a bilingual
standard terminology, containing fundus clinical terms and commonly used
descriptions in clinical diagnosis. Then, we establish two models,
RetSTA-7B-Zero and RetSTA-7B. RetSTA-7B-Zero, fine-tuned on an augmented
dataset simulating clinical scenarios, demonstrates powerful standardization
behaviors. However, it encounters a challenge of limitation to cover a wider
range of diseases. To further enhance standardization performance, we build
RetSTA-7B, which integrates a substantial amount of standardized data generated
by RetSTA-7B-Zero along with corresponding English data, covering diverse
complex clinical scenarios and achieving report-level standardization for the
first time. Experimental results demonstrate that RetSTA-7B outperforms other
compared LLMs in bilingual standardization task, which validates its superior
performance and generalizability. The checkpoints are available at
https://github.com/AB-Story/RetSTA-7B.
| no_new_dataset | 0.94079 |
2503.09361 | Katharina Prasse | Katharina Prasse, Marcel Kleinmann, Inken Adam, Kerstin
Beckersjuergen, Andreas Edte, Jona Frroku, Timotheus Gumpp, Steffen Jung,
Isaac Bravo, Stefanie Walter, Margret Keuper | Deep Learning for Climate Action: Computer Vision Analysis of Visual
Narratives on X | null | null | null | null | cs.CV cs.SI | http://creativecommons.org/licenses/by-sa/4.0/ | Climate change is one of the most pressing challenges of the 21st century,
sparking widespread discourse across social media platforms. Activists,
policymakers, and researchers seek to understand public sentiment and
narratives while access to social media data has become increasingly restricted
in the post-API era. In this study, we analyze a dataset of climate
change-related tweets from X (formerly Twitter) shared in 2019, containing 730k
tweets along with the shared images. Our approach integrates statistical
analysis, image classification, object detection, and sentiment analysis to
explore visual narratives in climate discourse. Additionally, we introduce a
graphical user interface (GUI) to facilitate interactive data exploration. Our
findings reveal key themes in climate communication, highlight sentiment
divergence between images and text, and underscore the strengths and
limitations of foundation models in analyzing social media imagery. By
releasing our code and tools, we aim to support future research on the
intersection of climate change, social media, and computer vision.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:03:49 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Prasse",
"Katharina",
""
],
[
"Kleinmann",
"Marcel",
""
],
[
"Adam",
"Inken",
""
],
[
"Beckersjuergen",
"Kerstin",
""
],
[
"Edte",
"Andreas",
""
],
[
"Frroku",
"Jona",
""
],
[
"Gumpp",
"Timotheus",
""
],
[
"Jung",
"Steffen",
""
],
[
"Bravo",
"Isaac",
""
],
[
"Walter",
"Stefanie",
""
],
[
"Keuper",
"Margret",
""
]
]
| TITLE: Deep Learning for Climate Action: Computer Vision Analysis of Visual
Narratives on X
ABSTRACT: Climate change is one of the most pressing challenges of the 21st century,
sparking widespread discourse across social media platforms. Activists,
policymakers, and researchers seek to understand public sentiment and
narratives while access to social media data has become increasingly restricted
in the post-API era. In this study, we analyze a dataset of climate
change-related tweets from X (formerly Twitter) shared in 2019, containing 730k
tweets along with the shared images. Our approach integrates statistical
analysis, image classification, object detection, and sentiment analysis to
explore visual narratives in climate discourse. Additionally, we introduce a
graphical user interface (GUI) to facilitate interactive data exploration. Our
findings reveal key themes in climate communication, highlight sentiment
divergence between images and text, and underscore the strengths and
limitations of foundation models in analyzing social media imagery. By
releasing our code and tools, we aim to support future research on the
intersection of climate change, social media, and computer vision.
| no_new_dataset | 0.878835 |
2503.09363 | Yuxiang Wang | Yuxiang Wang, Wenqi Fan, Suhang Wang, Yao Ma | Towards Graph Foundation Models: A Transferability Perspective | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, Graph Foundation Models (GFMs) have gained significant
attention for their potential to generalize across diverse graph domains and
tasks. Some works focus on Domain-Specific GFMs, which are designed to address
a variety of tasks within a specific domain, while others aim to create
General-Purpose GFMs that extend the capabilities of domain-specific models to
multiple domains. Regardless of the type, transferability is crucial for
applying GFMs across different domains and tasks. However, achieving strong
transferability is a major challenge due to the structural, feature, and
distributional variations in graph data. To date, there has been no systematic
research examining and analyzing GFMs from the perspective of transferability.
To bridge the gap, we present the first comprehensive taxonomy that categorizes
and analyzes existing GFMs through the lens of transferability, structuring
GFMs around their application scope (domain-specific vs. general-purpose) and
their approaches to knowledge acquisition and transfer. We provide a structured
perspective on current progress and identify potential pathways for advancing
GFM generalization across diverse graph datasets and tasks. We aims to shed
light on the current landscape of GFMs and inspire future research directions
in GFM development.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:04:05 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Wang",
"Yuxiang",
""
],
[
"Fan",
"Wenqi",
""
],
[
"Wang",
"Suhang",
""
],
[
"Ma",
"Yao",
""
]
]
| TITLE: Towards Graph Foundation Models: A Transferability Perspective
ABSTRACT: In recent years, Graph Foundation Models (GFMs) have gained significant
attention for their potential to generalize across diverse graph domains and
tasks. Some works focus on Domain-Specific GFMs, which are designed to address
a variety of tasks within a specific domain, while others aim to create
General-Purpose GFMs that extend the capabilities of domain-specific models to
multiple domains. Regardless of the type, transferability is crucial for
applying GFMs across different domains and tasks. However, achieving strong
transferability is a major challenge due to the structural, feature, and
distributional variations in graph data. To date, there has been no systematic
research examining and analyzing GFMs from the perspective of transferability.
To bridge the gap, we present the first comprehensive taxonomy that categorizes
and analyzes existing GFMs through the lens of transferability, structuring
GFMs around their application scope (domain-specific vs. general-purpose) and
their approaches to knowledge acquisition and transfer. We provide a structured
perspective on current progress and identify potential pathways for advancing
GFM generalization across diverse graph datasets and tasks. We aims to shed
light on the current landscape of GFMs and inspire future research directions
in GFM development.
| no_new_dataset | 0.939748 |
2503.09366 | Ziyi Huang | Ziyi Huang, Yang Li, Dushuai Li, Yao Mu, Hongmao Qin and Nan Zheng | Post-interactive Multimodal Trajectory Prediction for Autonomous Driving | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modeling the interactions among agents for trajectory prediction of
autonomous driving has been challenging due to the inherent uncertainty in
agents' behavior. The interactions involved in the predicted trajectories of
agents, also called post-interactions, have rarely been considered in
trajectory prediction models. To this end, we propose a coarse-to-fine
Transformer for multimodal trajectory prediction, i.e., Pioformer, which
explicitly extracts the post-interaction features to enhance the prediction
accuracy. Specifically, we first build a Coarse Trajectory Network to generate
coarse trajectories based on the observed trajectories and lane segments, in
which the low-order interaction features are extracted with the graph neural
networks. Next, we build a hypergraph neural network-based Trajectory Proposal
Network to generate trajectory proposals, where the high-order interaction
features are learned by the hypergraphs. Finally, the trajectory proposals are
sent to the Proposal Refinement Network for further refinement. The observed
trajectories and trajectory proposals are concatenated together as the inputs
of the Proposal Refinement Network, in which the post-interaction features are
learned by combining the previous interaction features and trajectory
consistency features. Moreover, we propose a three-stage training scheme to
facilitate the learning process. Extensive experiments on the Argoverse 1
dataset demonstrate the superiority of our method. Compared with the baseline
HiVT-64, our model has reduced the prediction errors by 4.4%, 8.4%, 14.4%, 5.7%
regarding metrics minADE6, minFDE6, MR6, and brier-minFDE6, respectively.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:10:09 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Huang",
"Ziyi",
""
],
[
"Li",
"Yang",
""
],
[
"Li",
"Dushuai",
""
],
[
"Mu",
"Yao",
""
],
[
"Qin",
"Hongmao",
""
],
[
"Zheng",
"Nan",
""
]
]
| TITLE: Post-interactive Multimodal Trajectory Prediction for Autonomous Driving
ABSTRACT: Modeling the interactions among agents for trajectory prediction of
autonomous driving has been challenging due to the inherent uncertainty in
agents' behavior. The interactions involved in the predicted trajectories of
agents, also called post-interactions, have rarely been considered in
trajectory prediction models. To this end, we propose a coarse-to-fine
Transformer for multimodal trajectory prediction, i.e., Pioformer, which
explicitly extracts the post-interaction features to enhance the prediction
accuracy. Specifically, we first build a Coarse Trajectory Network to generate
coarse trajectories based on the observed trajectories and lane segments, in
which the low-order interaction features are extracted with the graph neural
networks. Next, we build a hypergraph neural network-based Trajectory Proposal
Network to generate trajectory proposals, where the high-order interaction
features are learned by the hypergraphs. Finally, the trajectory proposals are
sent to the Proposal Refinement Network for further refinement. The observed
trajectories and trajectory proposals are concatenated together as the inputs
of the Proposal Refinement Network, in which the post-interaction features are
learned by combining the previous interaction features and trajectory
consistency features. Moreover, we propose a three-stage training scheme to
facilitate the learning process. Extensive experiments on the Argoverse 1
dataset demonstrate the superiority of our method. Compared with the baseline
HiVT-64, our model has reduced the prediction errors by 4.4%, 8.4%, 14.4%, 5.7%
regarding metrics minADE6, minFDE6, MR6, and brier-minFDE6, respectively.
| no_new_dataset | 0.951188 |
2503.09370 | Yang Nan | Yang Nan, Huichi Zhou, Xiaodan Xing, Giorgos Papanastasiou, Lei Zhu,
Zhifan Gao, Alejandro F Fangi, Guang Yang | Revisiting Medical Image Retrieval via Knowledge Consolidation | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As artificial intelligence and digital medicine increasingly permeate
healthcare systems, robust governance frameworks are essential to ensure
ethical, secure, and effective implementation. In this context, medical image
retrieval becomes a critical component of clinical data management, playing a
vital role in decision-making and safeguarding patient information. Existing
methods usually learn hash functions using bottleneck features, which fail to
produce representative hash codes from blended embeddings. Although contrastive
hashing has shown superior performance, current approaches often treat image
retrieval as a classification task, using category labels to create
positive/negative pairs. Moreover, many methods fail to address the
out-of-distribution (OOD) issue when models encounter external OOD queries or
adversarial attacks. In this work, we propose a novel method to consolidate
knowledge of hierarchical features and optimisation functions. We formulate the
knowledge consolidation by introducing Depth-aware Representation Fusion (DaRF)
and Structure-aware Contrastive Hashing (SCH). DaRF adaptively integrates
shallow and deep representations into blended features, and SCH incorporates
image fingerprints to enhance the adaptability of positive/negative pairings.
These blended features further facilitate OOD detection and content-based
recommendation, contributing to a secure AI-driven healthcare environment.
Moreover, we present a content-guided ranking to improve the robustness and
reproducibility of retrieval results. Our comprehensive assessments demonstrate
that the proposed method could effectively recognise OOD samples and
significantly outperform existing approaches in medical image retrieval
(p<0.05). In particular, our method achieves a 5.6-38.9% improvement in mean
Average Precision on the anatomical radiology dataset.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:16:42 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Nan",
"Yang",
""
],
[
"Zhou",
"Huichi",
""
],
[
"Xing",
"Xiaodan",
""
],
[
"Papanastasiou",
"Giorgos",
""
],
[
"Zhu",
"Lei",
""
],
[
"Gao",
"Zhifan",
""
],
[
"Fangi",
"Alejandro F",
""
],
[
"Yang",
"Guang",
""
]
]
| TITLE: Revisiting Medical Image Retrieval via Knowledge Consolidation
ABSTRACT: As artificial intelligence and digital medicine increasingly permeate
healthcare systems, robust governance frameworks are essential to ensure
ethical, secure, and effective implementation. In this context, medical image
retrieval becomes a critical component of clinical data management, playing a
vital role in decision-making and safeguarding patient information. Existing
methods usually learn hash functions using bottleneck features, which fail to
produce representative hash codes from blended embeddings. Although contrastive
hashing has shown superior performance, current approaches often treat image
retrieval as a classification task, using category labels to create
positive/negative pairs. Moreover, many methods fail to address the
out-of-distribution (OOD) issue when models encounter external OOD queries or
adversarial attacks. In this work, we propose a novel method to consolidate
knowledge of hierarchical features and optimisation functions. We formulate the
knowledge consolidation by introducing Depth-aware Representation Fusion (DaRF)
and Structure-aware Contrastive Hashing (SCH). DaRF adaptively integrates
shallow and deep representations into blended features, and SCH incorporates
image fingerprints to enhance the adaptability of positive/negative pairings.
These blended features further facilitate OOD detection and content-based
recommendation, contributing to a secure AI-driven healthcare environment.
Moreover, we present a content-guided ranking to improve the robustness and
reproducibility of retrieval results. Our comprehensive assessments demonstrate
that the proposed method could effectively recognise OOD samples and
significantly outperform existing approaches in medical image retrieval
(p<0.05). In particular, our method achieves a 5.6-38.9% improvement in mean
Average Precision on the anatomical radiology dataset.
| no_new_dataset | 0.9434 |
2503.09378 | Fangzheng Qi | Fangzheng Qi, Zhenjie Hou, En Lin, Xing Li, iuzhen Liang, Xinwen Zhou | Pig behavior dataset and Spatial-temporal perception and enhancement
networks based on the attention mechanism for pig behavior recognition | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recognition of pig behavior plays a crucial role in smart farming and
welfare assurance for pigs. Currently, in the field of pig behavior
recognition, the lack of publicly available behavioral datasets not only limits
the development of innovative algorithms but also hampers model robustness and
algorithm optimization.This paper proposes a dataset containing 13 pig
behaviors that significantly impact welfare.Based on this dataset, this paper
proposes a spatial-temporal perception and enhancement networks based on the
attention mechanism to model the spatiotemporal features of pig behaviors and
their associated interaction areas in video data. The network is composed of a
spatiotemporal perception network and a spatiotemporal feature enhancement
network. The spatiotemporal perception network is responsible for establishing
connections between the pigs and the key regions of their behaviors in the
video data. The spatiotemporal feature enhancement network further strengthens
the important spatial features of individual pigs and captures the long-term
dependencies of the spatiotemporal features of individual behaviors by
remodeling these connections, thereby enhancing the model's perception of
spatiotemporal changes in pig behaviors. Experimental results demonstrate that
on the dataset established in this paper, our proposed model achieves a MAP
score of 75.92%, which is an 8.17% improvement over the best-performing
traditional model. This study not only improces the accuracy and
generalizability of individual pig behavior recognition but also provides new
technological tools for modern smart farming. The dataset and related code will
be made publicly available alongside this paper.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:27:29 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Qi",
"Fangzheng",
""
],
[
"Hou",
"Zhenjie",
""
],
[
"Lin",
"En",
""
],
[
"Li",
"Xing",
""
],
[
"Liang",
"iuzhen",
""
],
[
"Zhou",
"Xinwen",
""
]
]
| TITLE: Pig behavior dataset and Spatial-temporal perception and enhancement
networks based on the attention mechanism for pig behavior recognition
ABSTRACT: The recognition of pig behavior plays a crucial role in smart farming and
welfare assurance for pigs. Currently, in the field of pig behavior
recognition, the lack of publicly available behavioral datasets not only limits
the development of innovative algorithms but also hampers model robustness and
algorithm optimization.This paper proposes a dataset containing 13 pig
behaviors that significantly impact welfare.Based on this dataset, this paper
proposes a spatial-temporal perception and enhancement networks based on the
attention mechanism to model the spatiotemporal features of pig behaviors and
their associated interaction areas in video data. The network is composed of a
spatiotemporal perception network and a spatiotemporal feature enhancement
network. The spatiotemporal perception network is responsible for establishing
connections between the pigs and the key regions of their behaviors in the
video data. The spatiotemporal feature enhancement network further strengthens
the important spatial features of individual pigs and captures the long-term
dependencies of the spatiotemporal features of individual behaviors by
remodeling these connections, thereby enhancing the model's perception of
spatiotemporal changes in pig behaviors. Experimental results demonstrate that
on the dataset established in this paper, our proposed model achieves a MAP
score of 75.92%, which is an 8.17% improvement over the best-performing
traditional model. This study not only improces the accuracy and
generalizability of individual pig behavior recognition but also provides new
technological tools for modern smart farming. The dataset and related code will
be made publicly available alongside this paper.
| new_dataset | 0.951908 |
2503.09382 | Jiani Huang | Jiani Huang, Shijie Wang, Liang-bo Ning, Wenqi Fan, Shuaiqiang Wang,
Dawei Yin, Qing Li | Towards Next-Generation Recommender Systems: A Benchmark for
Personalized Recommendation Assistant with LLMs | null | null | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recommender systems (RecSys) are widely used across various modern digital
platforms and have garnered significant attention. Traditional recommender
systems usually focus only on fixed and simple recommendation scenarios, making
it difficult to generalize to new and unseen recommendation tasks in an
interactive paradigm. Recently, the advancement of large language models (LLMs)
has revolutionized the foundational architecture of RecSys, driving their
evolution into more intelligent and interactive personalized recommendation
assistants. However, most existing studies rely on fixed task-specific prompt
templates to generate recommendations and evaluate the performance of
personalized assistants, which limits the comprehensive assessments of their
capabilities. This is because commonly used datasets lack high-quality textual
user queries that reflect real-world recommendation scenarios, making them
unsuitable for evaluating LLM-based personalized recommendation assistants. To
address this gap, we introduce RecBench+, a new dataset benchmark designed to
access LLMs' ability to handle intricate user recommendation needs in the era
of LLMs. RecBench+ encompasses a diverse set of queries that span both hard
conditions and soft preferences, with varying difficulty levels. We evaluated
commonly used LLMs on RecBench+ and uncovered below findings: 1) LLMs
demonstrate preliminary abilities to act as recommendation assistants, 2) LLMs
are better at handling queries with explicitly stated conditions, while facing
challenges with queries that require reasoning or contain misleading
information. Our dataset has been released at
https://github.com/jiani-huang/RecBench.git.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:28:23 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Huang",
"Jiani",
""
],
[
"Wang",
"Shijie",
""
],
[
"Ning",
"Liang-bo",
""
],
[
"Fan",
"Wenqi",
""
],
[
"Wang",
"Shuaiqiang",
""
],
[
"Yin",
"Dawei",
""
],
[
"Li",
"Qing",
""
]
]
| TITLE: Towards Next-Generation Recommender Systems: A Benchmark for
Personalized Recommendation Assistant with LLMs
ABSTRACT: Recommender systems (RecSys) are widely used across various modern digital
platforms and have garnered significant attention. Traditional recommender
systems usually focus only on fixed and simple recommendation scenarios, making
it difficult to generalize to new and unseen recommendation tasks in an
interactive paradigm. Recently, the advancement of large language models (LLMs)
has revolutionized the foundational architecture of RecSys, driving their
evolution into more intelligent and interactive personalized recommendation
assistants. However, most existing studies rely on fixed task-specific prompt
templates to generate recommendations and evaluate the performance of
personalized assistants, which limits the comprehensive assessments of their
capabilities. This is because commonly used datasets lack high-quality textual
user queries that reflect real-world recommendation scenarios, making them
unsuitable for evaluating LLM-based personalized recommendation assistants. To
address this gap, we introduce RecBench+, a new dataset benchmark designed to
access LLMs' ability to handle intricate user recommendation needs in the era
of LLMs. RecBench+ encompasses a diverse set of queries that span both hard
conditions and soft preferences, with varying difficulty levels. We evaluated
commonly used LLMs on RecBench+ and uncovered below findings: 1) LLMs
demonstrate preliminary abilities to act as recommendation assistants, 2) LLMs
are better at handling queries with explicitly stated conditions, while facing
challenges with queries that require reasoning or contain misleading
information. Our dataset has been released at
https://github.com/jiani-huang/RecBench.git.
| new_dataset | 0.964422 |
2503.09394 | Qiao Xiaozhen | Xiaozhen Qiao, Peng Huang, Jiakang Yuan, Xianda Guo, Bowen Ye, Zhe
Sun, Xuelong Li | Bidirectional Prototype-Reward co-Evolution for Test-Time Adaptation of
Vision-Language Models | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Test-time adaptation (TTA) is crucial in maintaining Vision-Language Models
(VLMs) performance when facing real-world distribution shifts, particularly
when the source data or target labels are inaccessible. Existing TTA methods
rely on CLIP's output probability distribution for feature evaluation, which
can introduce biases under domain shifts. This misalignment may cause features
to be misclassified due to text priors or incorrect textual associations. To
address these limitations, we propose Bidirectional Prototype-Reward
co-Evolution (BPRE), a novel TTA framework for VLMs that integrates feature
quality assessment with prototype evolution through a synergistic feedback
loop. BPRE first employs a Multi-Dimensional Quality-Aware Reward Module to
evaluate feature quality and guide prototype refinement precisely. The
continuous refinement of prototype quality through Prototype-Reward Interactive
Evolution will subsequently enhance the computation of more robust
Multi-Dimensional Quality-Aware Reward Scores. Through the bidirectional
interaction, the precision of rewards and the evolution of prototypes mutually
reinforce each other, forming a self-evolving cycle. Extensive experiments are
conducted across 15 diverse recognition datasets encompassing natural
distribution shifts and cross-dataset generalization scenarios. Results
demonstrate that BPRE consistently achieves superior average performance
compared to state-of-the-art methods across different model architectures, such
as ResNet-50 and ViT-B/16. By emphasizing comprehensive feature evaluation and
bidirectional knowledge refinement, BPRE advances VLM generalization
capabilities, offering a new perspective on TTA.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:40:33 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Qiao",
"Xiaozhen",
""
],
[
"Huang",
"Peng",
""
],
[
"Yuan",
"Jiakang",
""
],
[
"Guo",
"Xianda",
""
],
[
"Ye",
"Bowen",
""
],
[
"Sun",
"Zhe",
""
],
[
"Li",
"Xuelong",
""
]
]
| TITLE: Bidirectional Prototype-Reward co-Evolution for Test-Time Adaptation of
Vision-Language Models
ABSTRACT: Test-time adaptation (TTA) is crucial in maintaining Vision-Language Models
(VLMs) performance when facing real-world distribution shifts, particularly
when the source data or target labels are inaccessible. Existing TTA methods
rely on CLIP's output probability distribution for feature evaluation, which
can introduce biases under domain shifts. This misalignment may cause features
to be misclassified due to text priors or incorrect textual associations. To
address these limitations, we propose Bidirectional Prototype-Reward
co-Evolution (BPRE), a novel TTA framework for VLMs that integrates feature
quality assessment with prototype evolution through a synergistic feedback
loop. BPRE first employs a Multi-Dimensional Quality-Aware Reward Module to
evaluate feature quality and guide prototype refinement precisely. The
continuous refinement of prototype quality through Prototype-Reward Interactive
Evolution will subsequently enhance the computation of more robust
Multi-Dimensional Quality-Aware Reward Scores. Through the bidirectional
interaction, the precision of rewards and the evolution of prototypes mutually
reinforce each other, forming a self-evolving cycle. Extensive experiments are
conducted across 15 diverse recognition datasets encompassing natural
distribution shifts and cross-dataset generalization scenarios. Results
demonstrate that BPRE consistently achieves superior average performance
compared to state-of-the-art methods across different model architectures, such
as ResNet-50 and ViT-B/16. By emphasizing comprehensive feature evaluation and
bidirectional knowledge refinement, BPRE advances VLM generalization
capabilities, offering a new perspective on TTA.
| no_new_dataset | 0.950549 |
2503.09399 | Tobias Nauen | Tobias Christian Nauen, Brian Moser, Federico Raue, Stanislav Frolov,
Andreas Dengel | ForAug: Recombining Foregrounds and Backgrounds to Improve Vision
Transformer Training with Bias Mitigation | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transformers, particularly Vision Transformers (ViTs), have achieved
state-of-the-art performance in large-scale image classification. However, they
often require large amounts of data and can exhibit biases that limit their
robustness and generalizability. This paper introduces ForAug, a novel data
augmentation scheme that addresses these challenges and explicitly includes
inductive biases, which commonly are part of the neural network architecture,
into the training data. ForAug is constructed by using pretrained foundation
models to separate and recombine foreground objects with different backgrounds,
enabling fine-grained control over image composition during training. It thus
increases the data diversity and effective number of training samples. We
demonstrate that training on ForNet, the application of ForAug to ImageNet,
significantly improves the accuracy of ViTs and other architectures by up to
4.5 percentage points (p.p.) on ImageNet and 7.3 p.p. on downstream tasks.
Importantly, ForAug enables novel ways of analyzing model behavior and
quantifying biases. Namely, we introduce metrics for background robustness,
foreground focus, center bias, and size bias and show that training on ForNet
substantially reduces these biases compared to training on ImageNet. In
summary, ForAug provides a valuable tool for analyzing and mitigating biases,
enabling the development of more robust and reliable computer vision models.
Our code and dataset are publicly available at https://github.com/tobna/ForAug.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:49:45 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Nauen",
"Tobias Christian",
""
],
[
"Moser",
"Brian",
""
],
[
"Raue",
"Federico",
""
],
[
"Frolov",
"Stanislav",
""
],
[
"Dengel",
"Andreas",
""
]
]
| TITLE: ForAug: Recombining Foregrounds and Backgrounds to Improve Vision
Transformer Training with Bias Mitigation
ABSTRACT: Transformers, particularly Vision Transformers (ViTs), have achieved
state-of-the-art performance in large-scale image classification. However, they
often require large amounts of data and can exhibit biases that limit their
robustness and generalizability. This paper introduces ForAug, a novel data
augmentation scheme that addresses these challenges and explicitly includes
inductive biases, which commonly are part of the neural network architecture,
into the training data. ForAug is constructed by using pretrained foundation
models to separate and recombine foreground objects with different backgrounds,
enabling fine-grained control over image composition during training. It thus
increases the data diversity and effective number of training samples. We
demonstrate that training on ForNet, the application of ForAug to ImageNet,
significantly improves the accuracy of ViTs and other architectures by up to
4.5 percentage points (p.p.) on ImageNet and 7.3 p.p. on downstream tasks.
Importantly, ForAug enables novel ways of analyzing model behavior and
quantifying biases. Namely, we introduce metrics for background robustness,
foreground focus, center bias, and size bias and show that training on ForNet
substantially reduces these biases compared to training on ImageNet. In
summary, ForAug provides a valuable tool for analyzing and mitigating biases,
enabling the development of more robust and reliable computer vision models.
Our code and dataset are publicly available at https://github.com/tobna/ForAug.
| no_new_dataset | 0.943348 |
2503.09407 | Elaine Zosa | Elaine Zosa and Ville Komulainen and Sampo Pyysalo | Got Compute, but No Data: Lessons From Post-training a Finnish LLM | 7 pages | Proceedings of the Joint 25th Nordic Conference on Computational
Linguistics and 11th Baltic Conference on Human Language Technologies
(NoDaLiDa/Baltic-HLT 2025) | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As LLMs gain more popularity as chatbots and general assistants, methods have
been developed to enable LLMs to follow instructions and align with human
preferences. These methods have found success in the field, but their
effectiveness has not been demonstrated outside of high-resource languages. In
this work, we discuss our experiences in post-training an LLM for
instruction-following for English and Finnish. We use a multilingual LLM to
translate instruction and preference datasets from English to Finnish. We
perform instruction tuning and preference optimization in English and Finnish
and evaluate the instruction-following capabilities of the model in both
languages. Our results show that with a few hundred Finnish instruction samples
we can obtain competitive performance in Finnish instruction-following. We also
found that although preference optimization in English offers some
cross-lingual benefits, we obtain our best results by using preference data
from both languages. We release our model, datasets, and recipes under open
licenses at https://huggingface.co/LumiOpen/Poro-34B-chat-OpenAssistant
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:58:43 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zosa",
"Elaine",
""
],
[
"Komulainen",
"Ville",
""
],
[
"Pyysalo",
"Sampo",
""
]
]
| TITLE: Got Compute, but No Data: Lessons From Post-training a Finnish LLM
ABSTRACT: As LLMs gain more popularity as chatbots and general assistants, methods have
been developed to enable LLMs to follow instructions and align with human
preferences. These methods have found success in the field, but their
effectiveness has not been demonstrated outside of high-resource languages. In
this work, we discuss our experiences in post-training an LLM for
instruction-following for English and Finnish. We use a multilingual LLM to
translate instruction and preference datasets from English to Finnish. We
perform instruction tuning and preference optimization in English and Finnish
and evaluate the instruction-following capabilities of the model in both
languages. Our results show that with a few hundred Finnish instruction samples
we can obtain competitive performance in Finnish instruction-following. We also
found that although preference optimization in English offers some
cross-lingual benefits, we obtain our best results by using preference data
from both languages. We release our model, datasets, and recipes under open
licenses at https://huggingface.co/LumiOpen/Poro-34B-chat-OpenAssistant
| no_new_dataset | 0.634034 |
2503.09408 | Xiuzhen Guo | Xiuzhen Guo, Lianyuan Yu, Ji Shi, Na Lei, Hongxiao Wang | Diff-CL: A Novel Cross Pseudo-Supervision Method for Semi-supervised
Medical Image Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semi-supervised learning utilizes insights from unlabeled data to improve
model generalization, thereby reducing reliance on large labeled datasets. Most
existing studies focus on limited samples and fail to capture the overall data
distribution. We contend that combining distributional information with
detailed information is crucial for achieving more robust and accurate
segmentation results. On the one hand, with its robust generative capabilities,
diffusion models (DM) learn data distribution effectively. However, it
struggles with fine detail capture, leading to generated images with misleading
details. Combining DM with convolutional neural networks (CNNs) enables the
former to learn data distribution while the latter corrects fine details. While
capturing complete high-frequency details by CNNs requires substantial
computational resources and is susceptible to local noise. On the other hand,
given that both labeled and unlabeled data come from the same distribution, we
believe that regions in unlabeled data similar to overall class semantics to
labeled data are likely to belong to the same class, while regions with minimal
similarity are less likely to. This work introduces a semi-supervised medical
image segmentation framework from the distribution perspective (Diff-CL).
Firstly, we propose a cross-pseudo-supervision learning mechanism between
diffusion and convolution segmentation networks. Secondly, we design a
high-frequency mamba module to capture boundary and detail information
globally. Finally, we apply contrastive learning for label propagation from
labeled to unlabeled data. Our method achieves state-of-the-art (SOTA)
performance across three datasets, including left atrium, brain tumor, and NIH
pancreas datasets.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 13:59:09 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Guo",
"Xiuzhen",
""
],
[
"Yu",
"Lianyuan",
""
],
[
"Shi",
"Ji",
""
],
[
"Lei",
"Na",
""
],
[
"Wang",
"Hongxiao",
""
]
]
| TITLE: Diff-CL: A Novel Cross Pseudo-Supervision Method for Semi-supervised
Medical Image Segmentation
ABSTRACT: Semi-supervised learning utilizes insights from unlabeled data to improve
model generalization, thereby reducing reliance on large labeled datasets. Most
existing studies focus on limited samples and fail to capture the overall data
distribution. We contend that combining distributional information with
detailed information is crucial for achieving more robust and accurate
segmentation results. On the one hand, with its robust generative capabilities,
diffusion models (DM) learn data distribution effectively. However, it
struggles with fine detail capture, leading to generated images with misleading
details. Combining DM with convolutional neural networks (CNNs) enables the
former to learn data distribution while the latter corrects fine details. While
capturing complete high-frequency details by CNNs requires substantial
computational resources and is susceptible to local noise. On the other hand,
given that both labeled and unlabeled data come from the same distribution, we
believe that regions in unlabeled data similar to overall class semantics to
labeled data are likely to belong to the same class, while regions with minimal
similarity are less likely to. This work introduces a semi-supervised medical
image segmentation framework from the distribution perspective (Diff-CL).
Firstly, we propose a cross-pseudo-supervision learning mechanism between
diffusion and convolution segmentation networks. Secondly, we design a
high-frequency mamba module to capture boundary and detail information
globally. Finally, we apply contrastive learning for label propagation from
labeled to unlabeled data. Our method achieves state-of-the-art (SOTA)
performance across three datasets, including left atrium, brain tumor, and NIH
pancreas datasets.
| no_new_dataset | 0.953966 |
2503.09410 | Chen Zhao | Jiale Wang, Chen Zhao, Wei Ke, Tong Zhang | Monte Carlo Diffusion for Generalizable Learning-Based RANSAC | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Random Sample Consensus (RANSAC) is a fundamental approach for robustly
estimating parametric models from noisy data. Existing learning-based RANSAC
methods utilize deep learning to enhance the robustness of RANSAC against
outliers. However, these approaches are trained and tested on the data
generated by the same algorithms, leading to limited generalization to
out-of-distribution data during inference. Therefore, in this paper, we
introduce a novel diffusion-based paradigm that progressively injects noise
into ground-truth data, simulating the noisy conditions for training
learning-based RANSAC. To enhance data diversity, we incorporate Monte Carlo
sampling into the diffusion paradigm, approximating diverse data distributions
by introducing different types of randomness at multiple stages. We evaluate
our approach in the context of feature matching through comprehensive
experiments on the ScanNet and MegaDepth datasets. The experimental results
demonstrate that our Monte Carlo diffusion mechanism significantly improves the
generalization ability of learning-based RANSAC. We also develop extensive
ablation studies that highlight the effectiveness of key components in our
framework.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 14:01:18 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Wang",
"Jiale",
""
],
[
"Zhao",
"Chen",
""
],
[
"Ke",
"Wei",
""
],
[
"Zhang",
"Tong",
""
]
]
| TITLE: Monte Carlo Diffusion for Generalizable Learning-Based RANSAC
ABSTRACT: Random Sample Consensus (RANSAC) is a fundamental approach for robustly
estimating parametric models from noisy data. Existing learning-based RANSAC
methods utilize deep learning to enhance the robustness of RANSAC against
outliers. However, these approaches are trained and tested on the data
generated by the same algorithms, leading to limited generalization to
out-of-distribution data during inference. Therefore, in this paper, we
introduce a novel diffusion-based paradigm that progressively injects noise
into ground-truth data, simulating the noisy conditions for training
learning-based RANSAC. To enhance data diversity, we incorporate Monte Carlo
sampling into the diffusion paradigm, approximating diverse data distributions
by introducing different types of randomness at multiple stages. We evaluate
our approach in the context of feature matching through comprehensive
experiments on the ScanNet and MegaDepth datasets. The experimental results
demonstrate that our Monte Carlo diffusion mechanism significantly improves the
generalization ability of learning-based RANSAC. We also develop extensive
ablation studies that highlight the effectiveness of key components in our
framework.
| no_new_dataset | 0.948394 |
2503.09416 | WeiYing Xue | Qi Liu, Weiying Xue, Yuxiao Wang, Zhenao Wei | OpenVidVRD: Open-Vocabulary Video Visual Relation Detection via
Prompt-Driven Semantic Space Alignment | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The video visual relation detection (VidVRD) task is to identify objects and
their relationships in videos, which is challenging due to the dynamic content,
high annotation costs, and long-tailed distribution of relations. Visual
language models (VLMs) help explore open-vocabulary visual relation detection
tasks, yet often overlook the connections between various visual regions and
their relations. Moreover, using VLMs to directly identify visual relations in
videos poses significant challenges because of the large disparity between
images and videos. Therefore, we propose a novel open-vocabulary VidVRD
framework, termed OpenVidVRD, which transfers VLMs' rich knowledge and powerful
capabilities to improve VidVRD tasks through prompt learning. Specificall y, We
use VLM to extract text representations from automatically generated region
captions based on the video's regions. Next, we develop a spatiotemporal
refiner module to derive object-level relationship representations in the video
by integrating cross-modal spatiotemporal complementary information.
Furthermore, a prompt-driven strategy to align semantic spaces is employed to
harness the semantic understanding of VLMs, enhancing the overall
generalization ability of OpenVidVRD. Extensive experiments conducted on the
VidVRD and VidOR public datasets show that the proposed model outperforms
existing methods.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 14:13:17 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Liu",
"Qi",
""
],
[
"Xue",
"Weiying",
""
],
[
"Wang",
"Yuxiao",
""
],
[
"Wei",
"Zhenao",
""
]
]
| TITLE: OpenVidVRD: Open-Vocabulary Video Visual Relation Detection via
Prompt-Driven Semantic Space Alignment
ABSTRACT: The video visual relation detection (VidVRD) task is to identify objects and
their relationships in videos, which is challenging due to the dynamic content,
high annotation costs, and long-tailed distribution of relations. Visual
language models (VLMs) help explore open-vocabulary visual relation detection
tasks, yet often overlook the connections between various visual regions and
their relations. Moreover, using VLMs to directly identify visual relations in
videos poses significant challenges because of the large disparity between
images and videos. Therefore, we propose a novel open-vocabulary VidVRD
framework, termed OpenVidVRD, which transfers VLMs' rich knowledge and powerful
capabilities to improve VidVRD tasks through prompt learning. Specificall y, We
use VLM to extract text representations from automatically generated region
captions based on the video's regions. Next, we develop a spatiotemporal
refiner module to derive object-level relationship representations in the video
by integrating cross-modal spatiotemporal complementary information.
Furthermore, a prompt-driven strategy to align semantic spaces is employed to
harness the semantic understanding of VLMs, enhancing the overall
generalization ability of OpenVidVRD. Extensive experiments conducted on the
VidVRD and VidOR public datasets show that the proposed model outperforms
existing methods.
| no_new_dataset | 0.947137 |
2503.09427 | Yaorui Shi | Yaorui Shi, Jiaqi Yang, Sihang Li, Junfeng Fang, Xiang Wang, Zhiyuan
Liu, Yang Zhang | Multimodal Language Modeling for High-Accuracy Single Cell
Transcriptomics Analysis and Generation | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pre-trained language models (PLMs) have revolutionized scientific research,
yet their application to single-cell analysis remains limited. Text PLMs cannot
process single-cell RNA sequencing data, while cell PLMs lack the ability to
handle free text, restricting their use in multimodal tasks. Existing efforts
to bridge these modalities often suffer from information loss or inadequate
single-modal pre-training, leading to suboptimal performances. To address these
challenges, we propose Single-Cell MultiModal Generative Pre-trained
Transformer (scMMGPT), a unified PLM for joint cell and text modeling. scMMGPT
effectively integrates the state-of-the-art cell and text PLMs, facilitating
cross-modal knowledge sharing for improved performance. To bridge the text-cell
modality gap, scMMGPT leverages dedicated cross-modal projectors, and undergoes
extensive pre-training on 27 million cells -- the largest dataset for
multimodal cell-text PLMs to date. This large-scale pre-training enables
scMMGPT to excel in joint cell-text tasks, achieving an 84\% relative
improvement of textual discrepancy for cell description generation, 20.5\%
higher accuracy for cell type annotation, and 4\% improvement in $k$-NN
accuracy for text-conditioned pseudo-cell generation, outperforming baselines.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 14:26:16 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Shi",
"Yaorui",
""
],
[
"Yang",
"Jiaqi",
""
],
[
"Li",
"Sihang",
""
],
[
"Fang",
"Junfeng",
""
],
[
"Wang",
"Xiang",
""
],
[
"Liu",
"Zhiyuan",
""
],
[
"Zhang",
"Yang",
""
]
]
| TITLE: Multimodal Language Modeling for High-Accuracy Single Cell
Transcriptomics Analysis and Generation
ABSTRACT: Pre-trained language models (PLMs) have revolutionized scientific research,
yet their application to single-cell analysis remains limited. Text PLMs cannot
process single-cell RNA sequencing data, while cell PLMs lack the ability to
handle free text, restricting their use in multimodal tasks. Existing efforts
to bridge these modalities often suffer from information loss or inadequate
single-modal pre-training, leading to suboptimal performances. To address these
challenges, we propose Single-Cell MultiModal Generative Pre-trained
Transformer (scMMGPT), a unified PLM for joint cell and text modeling. scMMGPT
effectively integrates the state-of-the-art cell and text PLMs, facilitating
cross-modal knowledge sharing for improved performance. To bridge the text-cell
modality gap, scMMGPT leverages dedicated cross-modal projectors, and undergoes
extensive pre-training on 27 million cells -- the largest dataset for
multimodal cell-text PLMs to date. This large-scale pre-training enables
scMMGPT to excel in joint cell-text tasks, achieving an 84\% relative
improvement of textual discrepancy for cell description generation, 20.5\%
higher accuracy for cell type annotation, and 4\% improvement in $k$-NN
accuracy for text-conditioned pseudo-cell generation, outperforming baselines.
| no_new_dataset | 0.951051 |
2503.09439 | Qijian Zhang | Qijian Zhang, Xiaozheng Jian, Xuan Zhang, Wenping Wang, Junhui Hou | SuperCarver: Texture-Consistent 3D Geometry Super-Resolution for
High-Fidelity Surface Detail Generation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional production workflow of high-precision 3D mesh assets necessitates
a cumbersome and laborious process of manual sculpting by specialized modelers.
The recent years have witnessed remarkable advances in AI-empowered 3D content
creation. However, although the latest state-of-the-arts are already capable of
generating plausible structures and intricate appearances from images or text
prompts, the actual mesh surfaces are typically over-smoothing and lack
geometric details. This paper introduces SuperCarver, a 3D geometry
super-resolution framework particularly tailored for adding texture-consistent
surface details to given coarse meshes. Technically, we start by rendering the
original textured mesh into the image domain from multiple viewpoints. To
achieve geometric detail generation, we develop a deterministic prior-guided
normal diffusion model fine-tuned on a carefully curated dataset of paired
low-poly and high-poly normal renderings. To optimize mesh structures from
potentially imperfect normal map predictions, we design a simple yet effective
noise-resistant inverse rendering scheme based on distance field deformation.
Extensive experiments show that SuperCarver generates realistic and expressive
surface details as depicted by specific texture appearances, making it a
powerful tool for automatically upgrading massive outdated low-quality assets
and shortening the iteration cycle of high-quality mesh production in practical
applications.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 14:38:45 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Zhang",
"Qijian",
""
],
[
"Jian",
"Xiaozheng",
""
],
[
"Zhang",
"Xuan",
""
],
[
"Wang",
"Wenping",
""
],
[
"Hou",
"Junhui",
""
]
]
| TITLE: SuperCarver: Texture-Consistent 3D Geometry Super-Resolution for
High-Fidelity Surface Detail Generation
ABSTRACT: Traditional production workflow of high-precision 3D mesh assets necessitates
a cumbersome and laborious process of manual sculpting by specialized modelers.
The recent years have witnessed remarkable advances in AI-empowered 3D content
creation. However, although the latest state-of-the-arts are already capable of
generating plausible structures and intricate appearances from images or text
prompts, the actual mesh surfaces are typically over-smoothing and lack
geometric details. This paper introduces SuperCarver, a 3D geometry
super-resolution framework particularly tailored for adding texture-consistent
surface details to given coarse meshes. Technically, we start by rendering the
original textured mesh into the image domain from multiple viewpoints. To
achieve geometric detail generation, we develop a deterministic prior-guided
normal diffusion model fine-tuned on a carefully curated dataset of paired
low-poly and high-poly normal renderings. To optimize mesh structures from
potentially imperfect normal map predictions, we design a simple yet effective
noise-resistant inverse rendering scheme based on distance field deformation.
Extensive experiments show that SuperCarver generates realistic and expressive
surface details as depicted by specific texture appearances, making it a
powerful tool for automatically upgrading massive outdated low-quality assets
and shortening the iteration cycle of high-quality mesh production in practical
applications.
| new_dataset | 0.965119 |
2503.09443 | Julian Spravil | Julian Spravil and Sebastian Houben and Sven Behnke | Florenz: Scaling Laws for Systematic Generalization in Vision-Language
Models | null | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cross-lingual transfer enables vision-language models (VLMs) to perform
vision tasks in various languages with training data only in one language.
Current approaches rely on large pre-trained multilingual language models.
However, they face the curse of multilinguality, sacrificing downstream task
performance for multilingual capabilities, struggling with lexical ambiguities,
and falling behind recent advances. In this work, we study the scaling laws of
systematic generalization with monolingual VLMs for multilingual tasks,
focusing on the impact of model size and seen training samples. We propose
Florenz, a monolingual encoder-decoder VLM with 0.4B to 11.2B parameters
combining the pre-trained VLM Florence-2 and the large language model Gemma-2.
Florenz is trained with varying compute budgets on a synthetic dataset that
features intentionally incomplete language coverage for image captioning, thus,
testing generalization from the fully covered translation task. We show that
not only does indirectly learning unseen task-language pairs adhere to a
scaling law, but also that with our data generation pipeline and the proposed
Florenz model family, image captioning abilities can emerge in a specific
language even when only data for the translation task is available. Fine-tuning
on a mix of downstream datasets yields competitive performance and demonstrates
promising scaling trends in multimodal machine translation (Multi30K, CoMMuTE),
lexical disambiguation (CoMMuTE), and image captioning (Multi30K, XM3600, COCO
Karpathy).
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 14:41:10 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Spravil",
"Julian",
""
],
[
"Houben",
"Sebastian",
""
],
[
"Behnke",
"Sven",
""
]
]
| TITLE: Florenz: Scaling Laws for Systematic Generalization in Vision-Language
Models
ABSTRACT: Cross-lingual transfer enables vision-language models (VLMs) to perform
vision tasks in various languages with training data only in one language.
Current approaches rely on large pre-trained multilingual language models.
However, they face the curse of multilinguality, sacrificing downstream task
performance for multilingual capabilities, struggling with lexical ambiguities,
and falling behind recent advances. In this work, we study the scaling laws of
systematic generalization with monolingual VLMs for multilingual tasks,
focusing on the impact of model size and seen training samples. We propose
Florenz, a monolingual encoder-decoder VLM with 0.4B to 11.2B parameters
combining the pre-trained VLM Florence-2 and the large language model Gemma-2.
Florenz is trained with varying compute budgets on a synthetic dataset that
features intentionally incomplete language coverage for image captioning, thus,
testing generalization from the fully covered translation task. We show that
not only does indirectly learning unseen task-language pairs adhere to a
scaling law, but also that with our data generation pipeline and the proposed
Florenz model family, image captioning abilities can emerge in a specific
language even when only data for the translation task is available. Fine-tuning
on a mix of downstream datasets yields competitive performance and demonstrates
promising scaling trends in multimodal machine translation (Multi30K, CoMMuTE),
lexical disambiguation (CoMMuTE), and image captioning (Multi30K, XM3600, COCO
Karpathy).
| no_new_dataset | 0.947186 |
2503.09453 | Xiangjian Jiang | Xiangjian Jiang, Nikola Simidjievski, Mateja Jamnik | How Well Does Your Tabular Generator Learn the Structure of Tabular
Data? | Accepted by ICLR 2025 workshops (DeLTa and SynthData) | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Heterogeneous tabular data poses unique challenges in generative modelling
due to its fundamentally different underlying data structure compared to
homogeneous modalities, such as images and text. Although previous research has
sought to adapt the successes of generative modelling in homogeneous modalities
to the tabular domain, defining an effective generator for tabular data remains
an open problem. One major reason is that the evaluation criteria inherited
from other modalities often fail to adequately assess whether tabular
generative models effectively capture or utilise the unique structural
information encoded in tabular data. In this paper, we carefully examine the
limitations of the prevailing evaluation framework and introduce
$\textbf{TabStruct}$, a novel evaluation benchmark that positions structural
fidelity as a core evaluation dimension. Specifically, TabStruct evaluates the
alignment of causal structures in real and synthetic data, providing a direct
measure of how effectively tabular generative models learn the structure of
tabular data. Through extensive experiments using generators from eight
categories on seven datasets with expert-validated causal graphical structures,
we show that structural fidelity offers a task-independent, domain-agnostic
evaluation dimension. Our findings highlight the importance of tabular data
structure and offer practical guidance for developing more effective and robust
tabular generative models. Code is available at
https://github.com/SilenceX12138/TabStruct.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 14:54:58 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Jiang",
"Xiangjian",
""
],
[
"Simidjievski",
"Nikola",
""
],
[
"Jamnik",
"Mateja",
""
]
]
| TITLE: How Well Does Your Tabular Generator Learn the Structure of Tabular
Data?
ABSTRACT: Heterogeneous tabular data poses unique challenges in generative modelling
due to its fundamentally different underlying data structure compared to
homogeneous modalities, such as images and text. Although previous research has
sought to adapt the successes of generative modelling in homogeneous modalities
to the tabular domain, defining an effective generator for tabular data remains
an open problem. One major reason is that the evaluation criteria inherited
from other modalities often fail to adequately assess whether tabular
generative models effectively capture or utilise the unique structural
information encoded in tabular data. In this paper, we carefully examine the
limitations of the prevailing evaluation framework and introduce
$\textbf{TabStruct}$, a novel evaluation benchmark that positions structural
fidelity as a core evaluation dimension. Specifically, TabStruct evaluates the
alignment of causal structures in real and synthetic data, providing a direct
measure of how effectively tabular generative models learn the structure of
tabular data. Through extensive experiments using generators from eight
categories on seven datasets with expert-validated causal graphical structures,
we show that structural fidelity offers a task-independent, domain-agnostic
evaluation dimension. Our findings highlight the importance of tabular data
structure and offer practical guidance for developing more effective and robust
tabular generative models. Code is available at
https://github.com/SilenceX12138/TabStruct.
| no_new_dataset | 0.925567 |
2503.09474 | Mobarakol Islam | Jiayuan Huang, Runlong He, Danyal Z. Khan, Evangelos Mazomenos, Danail
Stoyanov, Hani J. Marcus, Matthew J. Clarkson, Mobarakol Islam | SurgicalVLM-Agent: Towards an Interactive AI Co-Pilot for Pituitary
Surgery | 11 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image-guided surgery demands adaptive, real-time decision support, yet static
AI models struggle with structured task planning and providing interactive
guidance. Large vision-language models (VLMs) offer a promising solution by
enabling dynamic task planning and predictive decision support. We introduce
SurgicalVLM-Agent, an AI co-pilot for image-guided pituitary surgery, capable
of conversation, planning, and task execution. The agent dynamically processes
surgeon queries and plans the tasks such as MRI tumor segmentation, endoscope
anatomy segmentation, overlaying preoperative imaging with intraoperative
views, instrument tracking, and surgical visual question answering (VQA). To
enable structured task planning, we develop the PitAgent dataset, a surgical
context-aware dataset covering segmentation, overlaying, instrument
localization, tool tracking, tool-tissue interactions, phase identification,
and surgical activity recognition. Additionally, we propose FFT-GaLore, a fast
Fourier transform (FFT)-based gradient projection technique for efficient
low-rank adaptation, optimizing fine-tuning for LLaMA 3.2 in surgical
environments. We validate SurgicalVLM-Agent by assessing task planning and
prompt generation on our PitAgent dataset and evaluating zero-shot VQA using a
public pituitary dataset. Results demonstrate state-of-the-art performance in
task planning and query interpretation, with highly semantically meaningful VQA
responses, advancing AI-driven surgical assistance.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 15:30:39 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Huang",
"Jiayuan",
""
],
[
"He",
"Runlong",
""
],
[
"Khan",
"Danyal Z.",
""
],
[
"Mazomenos",
"Evangelos",
""
],
[
"Stoyanov",
"Danail",
""
],
[
"Marcus",
"Hani J.",
""
],
[
"Clarkson",
"Matthew J.",
""
],
[
"Islam",
"Mobarakol",
""
]
]
| TITLE: SurgicalVLM-Agent: Towards an Interactive AI Co-Pilot for Pituitary
Surgery
ABSTRACT: Image-guided surgery demands adaptive, real-time decision support, yet static
AI models struggle with structured task planning and providing interactive
guidance. Large vision-language models (VLMs) offer a promising solution by
enabling dynamic task planning and predictive decision support. We introduce
SurgicalVLM-Agent, an AI co-pilot for image-guided pituitary surgery, capable
of conversation, planning, and task execution. The agent dynamically processes
surgeon queries and plans the tasks such as MRI tumor segmentation, endoscope
anatomy segmentation, overlaying preoperative imaging with intraoperative
views, instrument tracking, and surgical visual question answering (VQA). To
enable structured task planning, we develop the PitAgent dataset, a surgical
context-aware dataset covering segmentation, overlaying, instrument
localization, tool tracking, tool-tissue interactions, phase identification,
and surgical activity recognition. Additionally, we propose FFT-GaLore, a fast
Fourier transform (FFT)-based gradient projection technique for efficient
low-rank adaptation, optimizing fine-tuning for LLaMA 3.2 in surgical
environments. We validate SurgicalVLM-Agent by assessing task planning and
prompt generation on our PitAgent dataset and evaluating zero-shot VQA using a
public pituitary dataset. Results demonstrate state-of-the-art performance in
task planning and query interpretation, with highly semantically meaningful VQA
responses, advancing AI-driven surgical assistance.
| new_dataset | 0.961534 |
2503.09485 | Kadir Ozcoban | Kadir \"Oz\c{c}oban, Murat Manguo\u{g}lu, Emrullah Fatih Yetkin | A Novel Approach for Intrinsic Dimension Estimation | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The real-life data have a complex and non-linear structure due to their
nature. These non-linearities and the large number of features can usually
cause problems such as the empty-space phenomenon and the well-known curse of
dimensionality. Finding the nearly optimal representation of the dataset in a
lower-dimensional space (i.e. dimensionality reduction) offers an applicable
mechanism for improving the success of machine learning tasks. However,
estimating the required data dimension for the nearly optimal representation
(intrinsic dimension) can be very costly, particularly if one deals with big
data. We propose a highly efficient and robust intrinsic dimension estimation
approach that only relies on matrix-vector products for dimensionality
reduction methods. An experimental study is also conducted to compare the
performance of proposed method with state of the art approaches.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 15:42:39 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Özçoban",
"Kadir",
""
],
[
"Manguoğlu",
"Murat",
""
],
[
"Yetkin",
"Emrullah Fatih",
""
]
]
| TITLE: A Novel Approach for Intrinsic Dimension Estimation
ABSTRACT: The real-life data have a complex and non-linear structure due to their
nature. These non-linearities and the large number of features can usually
cause problems such as the empty-space phenomenon and the well-known curse of
dimensionality. Finding the nearly optimal representation of the dataset in a
lower-dimensional space (i.e. dimensionality reduction) offers an applicable
mechanism for improving the success of machine learning tasks. However,
estimating the required data dimension for the nearly optimal representation
(intrinsic dimension) can be very costly, particularly if one deals with big
data. We propose a highly efficient and robust intrinsic dimension estimation
approach that only relies on matrix-vector products for dimensionality
reduction methods. An experimental study is also conducted to compare the
performance of proposed method with state of the art approaches.
| no_new_dataset | 0.952175 |
2503.09493 | Romain Thoreau | Romain Thoreau, Valerio Marsocci and Dawa Derksen | Parameter-Efficient Adaptation of Geospatial Foundation Models through
Embedding Deflection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As large-scale heterogeneous data sets become increasingly available,
adapting foundation models at low cost has become a key issue. Seminal works in
natural language processing, e.g. Low-Rank Adaptation (LoRA), leverage the low
"intrinsic rank" of parameter updates during adaptation. In this paper, we
argue that incorporating stronger inductive biases in both data and models can
enhance the adaptation of Geospatial Foundation Models (GFMs), pretrained on
RGB satellite images, to other types of optical satellite data. Specifically,
the pretrained parameters of GFMs serve as a strong prior for the spatial
structure of multispectral images. For this reason, we introduce DEFLECT
(Deflecting Embeddings for Finetuning Latent representations for Earth and
Climate Tasks), a novel strategy for adapting GFMs to multispectral satellite
imagery with very few additional parameters. DEFLECT improves the
representation capabilities of the extracted features, particularly enhancing
spectral information, which is essential for geoscience and
environmental-related tasks. We demonstrate the effectiveness of our method
across three different GFMs and five diverse datasets, ranging from forest
monitoring to marine environment segmentation. Compared to competing methods,
DEFLECT achieves on-par or higher accuracy with 5-10$\times$ fewer parameters
for classification and segmentation tasks. The code will be made publicly
available.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 15:53:58 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Thoreau",
"Romain",
""
],
[
"Marsocci",
"Valerio",
""
],
[
"Derksen",
"Dawa",
""
]
]
| TITLE: Parameter-Efficient Adaptation of Geospatial Foundation Models through
Embedding Deflection
ABSTRACT: As large-scale heterogeneous data sets become increasingly available,
adapting foundation models at low cost has become a key issue. Seminal works in
natural language processing, e.g. Low-Rank Adaptation (LoRA), leverage the low
"intrinsic rank" of parameter updates during adaptation. In this paper, we
argue that incorporating stronger inductive biases in both data and models can
enhance the adaptation of Geospatial Foundation Models (GFMs), pretrained on
RGB satellite images, to other types of optical satellite data. Specifically,
the pretrained parameters of GFMs serve as a strong prior for the spatial
structure of multispectral images. For this reason, we introduce DEFLECT
(Deflecting Embeddings for Finetuning Latent representations for Earth and
Climate Tasks), a novel strategy for adapting GFMs to multispectral satellite
imagery with very few additional parameters. DEFLECT improves the
representation capabilities of the extracted features, particularly enhancing
spectral information, which is essential for geoscience and
environmental-related tasks. We demonstrate the effectiveness of our method
across three different GFMs and five diverse datasets, ranging from forest
monitoring to marine environment segmentation. Compared to competing methods,
DEFLECT achieves on-par or higher accuracy with 5-10$\times$ fewer parameters
for classification and segmentation tasks. The code will be made publicly
available.
| no_new_dataset | 0.951953 |
2503.09499 | Daoyuan Chen | Zhe Xu, Daoyuan Chen, Zhenqing Ling, Yaliang Li, Ying Shen | MindGYM: Enhancing Vision-Language Models via Synthetic Self-Challenging
Questions | 16 pages | null | null | null | cs.CV cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large vision-language models (VLMs) face challenges in achieving robust,
transferable reasoning abilities due to reliance on labor-intensive manual
instruction datasets or computationally expensive self-supervised methods. To
address these issues, we introduce MindGYM, a framework that enhances VLMs
through synthetic self-challenging questions, consisting of three stages: (1)
Seed Single-Hop Question Synthesis, generating cognitive questions across
textual (e.g., logical deduction) and multimodal contexts (e.g., diagram-based
queries) spanning eight semantic areas like ethical analysis; (2) Challenging
Multi-Hop Question Synthesis, combining seed questions via diverse principles
like bridging, visual-textual alignment, to create multi-step problems
demanding deeper reasoning; and (3) Thinking-Induced Curriculum Fine-Tuning, a
structured pipeline that progressively trains the model from scaffolded
reasoning to standalone inference. By leveraging the model's self-synthesis
capability, MindGYM achieves high data efficiency (e.g., +16% gains on
MathVision-Mini with only 400 samples), computational efficiency (reducing both
training and inference costs), and robust generalization across tasks.
Extensive evaluations on seven benchmarks demonstrate superior performance over
strong baselines, with notable improvements (+15.77% win rates) in reasoning
depth and breadth validated via GPT-based scoring. MindGYM underscores the
viability of self-challenging for refining VLM capabilities while minimizing
human intervention and resource demands. Code and data are released to advance
multimodal reasoning research.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:03:03 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Xu",
"Zhe",
""
],
[
"Chen",
"Daoyuan",
""
],
[
"Ling",
"Zhenqing",
""
],
[
"Li",
"Yaliang",
""
],
[
"Shen",
"Ying",
""
]
]
| TITLE: MindGYM: Enhancing Vision-Language Models via Synthetic Self-Challenging
Questions
ABSTRACT: Large vision-language models (VLMs) face challenges in achieving robust,
transferable reasoning abilities due to reliance on labor-intensive manual
instruction datasets or computationally expensive self-supervised methods. To
address these issues, we introduce MindGYM, a framework that enhances VLMs
through synthetic self-challenging questions, consisting of three stages: (1)
Seed Single-Hop Question Synthesis, generating cognitive questions across
textual (e.g., logical deduction) and multimodal contexts (e.g., diagram-based
queries) spanning eight semantic areas like ethical analysis; (2) Challenging
Multi-Hop Question Synthesis, combining seed questions via diverse principles
like bridging, visual-textual alignment, to create multi-step problems
demanding deeper reasoning; and (3) Thinking-Induced Curriculum Fine-Tuning, a
structured pipeline that progressively trains the model from scaffolded
reasoning to standalone inference. By leveraging the model's self-synthesis
capability, MindGYM achieves high data efficiency (e.g., +16% gains on
MathVision-Mini with only 400 samples), computational efficiency (reducing both
training and inference costs), and robust generalization across tasks.
Extensive evaluations on seven benchmarks demonstrate superior performance over
strong baselines, with notable improvements (+15.77% win rates) in reasoning
depth and breadth validated via GPT-based scoring. MindGYM underscores the
viability of self-challenging for refining VLM capabilities while minimizing
human intervention and resource demands. Code and data are released to advance
multimodal reasoning research.
| no_new_dataset | 0.949482 |
2503.09504 | Bakary Badjie Mr | Bakary Badjie, Jos\'e Cec\'ilio, Ant\'onio Casimiro | Double-Stage Feature-Level Clustering-Based Mixture of Experts Framework | 14 Pages, 1 Figure, and 3 Tables | null | null | null | cs.LG cs.AI cs.CV cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Mixture-of-Experts (MoE) model has succeeded in deep learning (DL).
However, its complex architecture and advantages over dense models in image
classification remain unclear. In previous studies, MoE performance has often
been affected by noise and outliers in the input space. Some approaches
incorporate input clustering for training MoE models, but most clustering
algorithms lack access to labeled data, limiting their effectiveness. This
paper introduces the Double-stage Feature-level Clustering and
Pseudo-labeling-based Mixture of Experts (DFCP-MoE) framework, which consists
of input feature extraction, feature-level clustering, and a computationally
efficient pseudo-labeling strategy. This approach reduces the impact of noise
and outliers while leveraging a small subset of labeled data to label a large
portion of unlabeled inputs. We propose a conditional end-to-end joint training
method that improves expert specialization by training the MoE model on
well-labeled, clustered inputs. Unlike traditional MoE and dense models, the
DFCP-MoE framework effectively captures input space diversity, leading to
competitive inference results. We validate our approach on three benchmark
datasets for multi-class classification tasks.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:13:50 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Badjie",
"Bakary",
""
],
[
"Cecílio",
"José",
""
],
[
"Casimiro",
"António",
""
]
]
| TITLE: Double-Stage Feature-Level Clustering-Based Mixture of Experts Framework
ABSTRACT: The Mixture-of-Experts (MoE) model has succeeded in deep learning (DL).
However, its complex architecture and advantages over dense models in image
classification remain unclear. In previous studies, MoE performance has often
been affected by noise and outliers in the input space. Some approaches
incorporate input clustering for training MoE models, but most clustering
algorithms lack access to labeled data, limiting their effectiveness. This
paper introduces the Double-stage Feature-level Clustering and
Pseudo-labeling-based Mixture of Experts (DFCP-MoE) framework, which consists
of input feature extraction, feature-level clustering, and a computationally
efficient pseudo-labeling strategy. This approach reduces the impact of noise
and outliers while leveraging a small subset of labeled data to label a large
portion of unlabeled inputs. We propose a conditional end-to-end joint training
method that improves expert specialization by training the MoE model on
well-labeled, clustered inputs. Unlike traditional MoE and dense models, the
DFCP-MoE framework effectively captures input space diversity, leading to
competitive inference results. We validate our approach on three benchmark
datasets for multi-class classification tasks.
| no_new_dataset | 0.950641 |
2503.09510 | Rosalia Tufano | Rosalia Tufano, Gabriele Bavota | Automating Code Review: A Systematic Literature Review | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Code Review consists in assessing the code written by teammates with the goal
of increasing code quality. Empirical studies documented the benefits brought
by such a practice that, however, has its cost to pay in terms of developers'
time. For this reason, researchers have proposed techniques and tools to
automate code review tasks such as the reviewers selection (i.e., identifying
suitable reviewers for a given code change) or the actual review of a given
change (i.e., recommending improvements to the contributor as a human reviewer
would do). Given the substantial amount of papers recently published on the
topic, it may be challenging for researchers and practitioners to get a
complete overview of the state-of-the-art.
We present a systematic literature review (SLR) featuring 119 papers
concerning the automation of code review tasks. We provide: (i) a
categorization of the code review tasks automated in the literature; (ii) an
overview of the under-the-hood techniques used for the automation, including
the datasets used for training data-driven techniques; (iii) publicly available
techniques and datasets used for their evaluation, with a description of the
evaluation metrics usually adopted for each task.
The SLR is concluded by a discussion of the current limitations of the
state-of-the-art, with insights for future research directions.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:19:10 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Tufano",
"Rosalia",
""
],
[
"Bavota",
"Gabriele",
""
]
]
| TITLE: Automating Code Review: A Systematic Literature Review
ABSTRACT: Code Review consists in assessing the code written by teammates with the goal
of increasing code quality. Empirical studies documented the benefits brought
by such a practice that, however, has its cost to pay in terms of developers'
time. For this reason, researchers have proposed techniques and tools to
automate code review tasks such as the reviewers selection (i.e., identifying
suitable reviewers for a given code change) or the actual review of a given
change (i.e., recommending improvements to the contributor as a human reviewer
would do). Given the substantial amount of papers recently published on the
topic, it may be challenging for researchers and practitioners to get a
complete overview of the state-of-the-art.
We present a systematic literature review (SLR) featuring 119 papers
concerning the automation of code review tasks. We provide: (i) a
categorization of the code review tasks automated in the literature; (ii) an
overview of the under-the-hood techniques used for the automation, including
the datasets used for training data-driven techniques; (iii) publicly available
techniques and datasets used for their evaluation, with a description of the
evaluation metrics usually adopted for each task.
The SLR is concluded by a discussion of the current limitations of the
state-of-the-art, with insights for future research directions.
| no_new_dataset | 0.946892 |
2503.09514 | Bin Hu | Bin Hu and Chenqiang Gao and Shurui Liu and Junjie Guo and Fang Chen
and Fangcen Liu | CM-Diff: A Single Generative Network for Bidirectional Cross-Modality
Translation Diffusion Model Between Infrared and Visible Images | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The image translation method represents a crucial approach for mitigating
information deficiencies in the infrared and visible modalities, while also
facilitating the enhancement of modality-specific datasets. However, existing
methods for infrared and visible image translation either achieve
unidirectional modality translation or rely on cycle consistency for
bidirectional modality translation, which may result in suboptimal performance.
In this work, we present the cross-modality translation diffusion model
(CM-Diff) for simultaneously modeling data distributions in both the infrared
and visible modalities. We address this challenge by combining translation
direction labels for guidance during training with cross-modality feature
control. Specifically, we view the establishment of the mapping relationship
between the two modalities as the process of learning data distributions and
understanding modality differences, achieved through a novel Bidirectional
Diffusion Training (BDT) strategy. Additionally, we propose a Statistical
Constraint Inference (SCI) strategy to ensure the generated image closely
adheres to the data distribution of the target modality. Experimental results
demonstrate the superiority of our CM-Diff over state-of-the-art methods,
highlighting its potential for generating dual-modality datasets.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:25:18 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Hu",
"Bin",
""
],
[
"Gao",
"Chenqiang",
""
],
[
"Liu",
"Shurui",
""
],
[
"Guo",
"Junjie",
""
],
[
"Chen",
"Fang",
""
],
[
"Liu",
"Fangcen",
""
]
]
| TITLE: CM-Diff: A Single Generative Network for Bidirectional Cross-Modality
Translation Diffusion Model Between Infrared and Visible Images
ABSTRACT: The image translation method represents a crucial approach for mitigating
information deficiencies in the infrared and visible modalities, while also
facilitating the enhancement of modality-specific datasets. However, existing
methods for infrared and visible image translation either achieve
unidirectional modality translation or rely on cycle consistency for
bidirectional modality translation, which may result in suboptimal performance.
In this work, we present the cross-modality translation diffusion model
(CM-Diff) for simultaneously modeling data distributions in both the infrared
and visible modalities. We address this challenge by combining translation
direction labels for guidance during training with cross-modality feature
control. Specifically, we view the establishment of the mapping relationship
between the two modalities as the process of learning data distributions and
understanding modality differences, achieved through a novel Bidirectional
Diffusion Training (BDT) strategy. Additionally, we propose a Statistical
Constraint Inference (SCI) strategy to ensure the generated image closely
adheres to the data distribution of the target modality. Experimental results
demonstrate the superiority of our CM-Diff over state-of-the-art methods,
highlighting its potential for generating dual-modality datasets.
| no_new_dataset | 0.95253 |
2503.09527 | Peng Chen | Peng Chen, Pi Bu, Yingyao Wang, Xinyi Wang, Ziming Wang, Jie Guo,
Yingxiu Zhao, Qi Zhu, Jun Song, Siran Yang, Jiamang Wang, Bo Zheng | CombatVLA: An Efficient Vision-Language-Action Model for Combat Tasks in
3D Action Role-Playing Games | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recent advances in Vision-Language-Action models (VLAs) have expanded the
capabilities of embodied intelligence. However, significant challenges remain
in real-time decision-making in complex 3D environments, which demand
second-level responses, high-resolution perception, and tactical reasoning
under dynamic conditions. To advance the field, we introduce CombatVLA, an
efficient VLA model optimized for combat tasks in 3D action role-playing
games(ARPGs). Specifically, our CombatVLA is a 3B model trained on video-action
pairs collected by an action tracker, where the data is formatted as
action-of-thought (AoT) sequences. Thereafter, CombatVLA seamlessly integrates
into an action execution framework, allowing efficient inference through our
truncated AoT strategy. Experimental results demonstrate that CombatVLA not
only outperforms all existing models on the combat understanding benchmark but
also achieves a 50-fold acceleration in game combat. Moreover, it has a higher
task success rate than human players. We will open-source all resources,
including the action tracker, dataset, benchmark, model weights, training code,
and the implementation of the framework at https://combatvla.github.io/.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:42:26 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Chen",
"Peng",
""
],
[
"Bu",
"Pi",
""
],
[
"Wang",
"Yingyao",
""
],
[
"Wang",
"Xinyi",
""
],
[
"Wang",
"Ziming",
""
],
[
"Guo",
"Jie",
""
],
[
"Zhao",
"Yingxiu",
""
],
[
"Zhu",
"Qi",
""
],
[
"Song",
"Jun",
""
],
[
"Yang",
"Siran",
""
],
[
"Wang",
"Jiamang",
""
],
[
"Zheng",
"Bo",
""
]
]
| TITLE: CombatVLA: An Efficient Vision-Language-Action Model for Combat Tasks in
3D Action Role-Playing Games
ABSTRACT: Recent advances in Vision-Language-Action models (VLAs) have expanded the
capabilities of embodied intelligence. However, significant challenges remain
in real-time decision-making in complex 3D environments, which demand
second-level responses, high-resolution perception, and tactical reasoning
under dynamic conditions. To advance the field, we introduce CombatVLA, an
efficient VLA model optimized for combat tasks in 3D action role-playing
games(ARPGs). Specifically, our CombatVLA is a 3B model trained on video-action
pairs collected by an action tracker, where the data is formatted as
action-of-thought (AoT) sequences. Thereafter, CombatVLA seamlessly integrates
into an action execution framework, allowing efficient inference through our
truncated AoT strategy. Experimental results demonstrate that CombatVLA not
only outperforms all existing models on the combat understanding benchmark but
also achieves a 50-fold acceleration in game combat. Moreover, it has a higher
task success rate than human players. We will open-source all resources,
including the action tracker, dataset, benchmark, model weights, training code,
and the implementation of the framework at https://combatvla.github.io/.
| no_new_dataset | 0.940735 |
2503.09535 | Utku Ozbulak | Minjae Chung, Jong Bum Won, Ganghyun Kim, Yujin Kim, Utku Ozbulak | Evaluating Visual Explanations of Attention Maps for Transformer-based
Medical Imaging | Accepted for publication in MICCAI 2024 Workshop on Interpretability
of Machine Intelligence in Medical Image Computing (iMIMIC) | null | 10.1007/978-3-031-77610-6_11 | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although Vision Transformers (ViTs) have recently demonstrated superior
performance in medical imaging problems, they face explainability issues
similar to previous architectures such as convolutional neural networks. Recent
research efforts suggest that attention maps, which are part of decision-making
process of ViTs can potentially address the explainability issue by identifying
regions influencing predictions, especially in models pretrained with
self-supervised learning. In this work, we compare the visual explanations of
attention maps to other commonly used methods for medical imaging problems. To
do so, we employ four distinct medical imaging datasets that involve the
identification of (1) colonic polyps, (2) breast tumors, (3) esophageal
inflammation, and (4) bone fractures and hardware implants. Through large-scale
experiments on the aforementioned datasets using various supervised and
self-supervised pretrained ViTs, we find that although attention maps show
promise under certain conditions and generally surpass GradCAM in
explainability, they are outperformed by transformer-specific interpretability
methods. Our findings indicate that the efficacy of attention maps as a method
of interpretability is context-dependent and may be limited as they do not
consistently provide the comprehensive insights required for robust medical
decision-making.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:52:52 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Chung",
"Minjae",
""
],
[
"Won",
"Jong Bum",
""
],
[
"Kim",
"Ganghyun",
""
],
[
"Kim",
"Yujin",
""
],
[
"Ozbulak",
"Utku",
""
]
]
| TITLE: Evaluating Visual Explanations of Attention Maps for Transformer-based
Medical Imaging
ABSTRACT: Although Vision Transformers (ViTs) have recently demonstrated superior
performance in medical imaging problems, they face explainability issues
similar to previous architectures such as convolutional neural networks. Recent
research efforts suggest that attention maps, which are part of decision-making
process of ViTs can potentially address the explainability issue by identifying
regions influencing predictions, especially in models pretrained with
self-supervised learning. In this work, we compare the visual explanations of
attention maps to other commonly used methods for medical imaging problems. To
do so, we employ four distinct medical imaging datasets that involve the
identification of (1) colonic polyps, (2) breast tumors, (3) esophageal
inflammation, and (4) bone fractures and hardware implants. Through large-scale
experiments on the aforementioned datasets using various supervised and
self-supervised pretrained ViTs, we find that although attention maps show
promise under certain conditions and generally surpass GradCAM in
explainability, they are outperformed by transformer-specific interpretability
methods. Our findings indicate that the efficacy of attention maps as a method
of interpretability is context-dependent and may be limited as they do not
consistently provide the comprehensive insights required for robust medical
decision-making.
| no_new_dataset | 0.945851 |
2503.09537 | Shuokang Huang | Shuokang Huang, Julie A. McCann | GenHPE: Generative Counterfactuals for 3D Human Pose Estimation with
Radio Frequency Signals | null | null | null | null | cs.CV cs.AI cs.MM eess.SP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Human pose estimation (HPE) detects the positions of human body joints for
various applications. Compared to using cameras, HPE using radio frequency (RF)
signals is non-intrusive and more robust to adverse conditions, exploiting the
signal variations caused by human interference. However, existing studies focus
on single-domain HPE confined by domain-specific confounders, which cannot
generalize to new domains and result in diminished HPE performance.
Specifically, the signal variations caused by different human body parts are
entangled, containing subject-specific confounders. RF signals are also
intertwined with environmental noise, involving environment-specific
confounders. In this paper, we propose GenHPE, a 3D HPE approach that generates
counterfactual RF signals to eliminate domain-specific confounders. GenHPE
trains generative models conditioned on human skeleton labels, learning how
human body parts and confounders interfere with RF signals. We manipulate
skeleton labels (i.e., removing body parts) as counterfactual conditions for
generative models to synthesize counterfactual RF signals. The differences
between counterfactual signals approximately eliminate domain-specific
confounders and regularize an encoder-decoder model to learn domain-independent
representations. Such representations help GenHPE generalize to new
subjects/environments for cross-domain 3D HPE. We evaluate GenHPE on three
public datasets from WiFi, ultra-wideband, and millimeter wave. Experimental
results show that GenHPE outperforms state-of-the-art methods and reduces
estimation errors by up to 52.2mm for cross-subject HPE and 10.6mm for
cross-environment HPE.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 16:53:58 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Huang",
"Shuokang",
""
],
[
"McCann",
"Julie A.",
""
]
]
| TITLE: GenHPE: Generative Counterfactuals for 3D Human Pose Estimation with
Radio Frequency Signals
ABSTRACT: Human pose estimation (HPE) detects the positions of human body joints for
various applications. Compared to using cameras, HPE using radio frequency (RF)
signals is non-intrusive and more robust to adverse conditions, exploiting the
signal variations caused by human interference. However, existing studies focus
on single-domain HPE confined by domain-specific confounders, which cannot
generalize to new domains and result in diminished HPE performance.
Specifically, the signal variations caused by different human body parts are
entangled, containing subject-specific confounders. RF signals are also
intertwined with environmental noise, involving environment-specific
confounders. In this paper, we propose GenHPE, a 3D HPE approach that generates
counterfactual RF signals to eliminate domain-specific confounders. GenHPE
trains generative models conditioned on human skeleton labels, learning how
human body parts and confounders interfere with RF signals. We manipulate
skeleton labels (i.e., removing body parts) as counterfactual conditions for
generative models to synthesize counterfactual RF signals. The differences
between counterfactual signals approximately eliminate domain-specific
confounders and regularize an encoder-decoder model to learn domain-independent
representations. Such representations help GenHPE generalize to new
subjects/environments for cross-domain 3D HPE. We evaluate GenHPE on three
public datasets from WiFi, ultra-wideband, and millimeter wave. Experimental
results show that GenHPE outperforms state-of-the-art methods and reduces
estimation errors by up to 52.2mm for cross-subject HPE and 10.6mm for
cross-environment HPE.
| no_new_dataset | 0.946597 |
2503.09551 | Christopher Leon | Christopher Leon, Petr M. Anisimov, Nikolai Yampolsky, Alexander
Scheinker | Using Convolutional Neural Networks to Accelerate 3D Coherent
Synchrotron Radiation Computations | null | null | null | null | physics.acc-ph | http://creativecommons.org/licenses/by/4.0/ | Calculating the effects of Coherent Synchrotron Radiation (CSR) is one of the
most computationally expensive tasks in accelerator physics. Here, we use
convolutional neural networks (CNN's), along with a latent conditional
diffusion (LCD) model, trained on physics-based simulations to speed up
calculations. Specifically, we produce the 3D CSR wakefields generated by
electron bunches in circular orbit in the steady-state condition. Two datasets
are used for training and testing the models: wakefields generated by
three-dimensional Gaussian electron distributions and wakefields from a sum of
up to 25 three-dimensional Gaussian distributions. The CNN's are able to
accurately produce the 3D wakefields $\sim 250-1000$ times faster than the
numerical calculations, while the LCD has a gain of a factor of $\sim 34$. We
also test the extrapolation and out-of-distribution generalization ability of
the models. They generalize well on distributions with larger spreads than what
they were trained on, but struggle with smaller spreads.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 17:08:13 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Leon",
"Christopher",
""
],
[
"Anisimov",
"Petr M.",
""
],
[
"Yampolsky",
"Nikolai",
""
],
[
"Scheinker",
"Alexander",
""
]
]
| TITLE: Using Convolutional Neural Networks to Accelerate 3D Coherent
Synchrotron Radiation Computations
ABSTRACT: Calculating the effects of Coherent Synchrotron Radiation (CSR) is one of the
most computationally expensive tasks in accelerator physics. Here, we use
convolutional neural networks (CNN's), along with a latent conditional
diffusion (LCD) model, trained on physics-based simulations to speed up
calculations. Specifically, we produce the 3D CSR wakefields generated by
electron bunches in circular orbit in the steady-state condition. Two datasets
are used for training and testing the models: wakefields generated by
three-dimensional Gaussian electron distributions and wakefields from a sum of
up to 25 three-dimensional Gaussian distributions. The CNN's are able to
accurately produce the 3D wakefields $\sim 250-1000$ times faster than the
numerical calculations, while the LCD has a gain of a factor of $\sim 34$. We
also test the extrapolation and out-of-distribution generalization ability of
the models. They generalize well on distributions with larger spreads than what
they were trained on, but struggle with smaller spreads.
| no_new_dataset | 0.948106 |
2503.09556 | Tim B\"uchner | Tim B\"uchner, Christoph Anders, Orlando Guntinas-Lichius, Joachim
Denzler | Electromyography-Informed Facial Expression Reconstruction for
Physiological-Based Synthesis and Analysis | Accepted at CVPR 2025, 41 pages, 37 figures, 8 tables | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The relationship between muscle activity and resulting facial expressions is
crucial for various fields, including psychology, medicine, and entertainment.
The synchronous recording of facial mimicry and muscular activity via surface
electromyography (sEMG) provides a unique window into these complex dynamics.
Unfortunately, existing methods for facial analysis cannot handle electrode
occlusion, rendering them ineffective. Even with occlusion-free reference
images of the same person, variations in expression intensity and execution are
unmatchable. Our electromyography-informed facial expression reconstruction
(EIFER) approach is a novel method to restore faces under sEMG occlusion
faithfully in an adversarial manner. We decouple facial geometry and visual
appearance (e.g., skin texture, lighting, electrodes) by combining a 3D
Morphable Model (3DMM) with neural unpaired image-to-image translation via
reference recordings. Then, EIFER learns a bidirectional mapping between 3DMM
expression parameters and muscle activity, establishing correspondence between
the two domains. We validate the effectiveness of our approach through
experiments on a dataset of synchronized sEMG recordings and facial mimicry,
demonstrating faithful geometry and appearance reconstruction. Further, we
synthesize expressions based on muscle activity and how observed expressions
can predict dynamic muscle activity. Consequently, EIFER introduces a new
paradigm for facial electromyography, which could be extended to other forms of
multi-modal face recordings.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 17:21:10 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Büchner",
"Tim",
""
],
[
"Anders",
"Christoph",
""
],
[
"Guntinas-Lichius",
"Orlando",
""
],
[
"Denzler",
"Joachim",
""
]
]
| TITLE: Electromyography-Informed Facial Expression Reconstruction for
Physiological-Based Synthesis and Analysis
ABSTRACT: The relationship between muscle activity and resulting facial expressions is
crucial for various fields, including psychology, medicine, and entertainment.
The synchronous recording of facial mimicry and muscular activity via surface
electromyography (sEMG) provides a unique window into these complex dynamics.
Unfortunately, existing methods for facial analysis cannot handle electrode
occlusion, rendering them ineffective. Even with occlusion-free reference
images of the same person, variations in expression intensity and execution are
unmatchable. Our electromyography-informed facial expression reconstruction
(EIFER) approach is a novel method to restore faces under sEMG occlusion
faithfully in an adversarial manner. We decouple facial geometry and visual
appearance (e.g., skin texture, lighting, electrodes) by combining a 3D
Morphable Model (3DMM) with neural unpaired image-to-image translation via
reference recordings. Then, EIFER learns a bidirectional mapping between 3DMM
expression parameters and muscle activity, establishing correspondence between
the two domains. We validate the effectiveness of our approach through
experiments on a dataset of synchronized sEMG recordings and facial mimicry,
demonstrating faithful geometry and appearance reconstruction. Further, we
synthesize expressions based on muscle activity and how observed expressions
can predict dynamic muscle activity. Consequently, EIFER introduces a new
paradigm for facial electromyography, which could be extended to other forms of
multi-modal face recordings.
| no_new_dataset | 0.850282 |
2503.09565 | Zixiang Chen | Zixiang Chen, Greg Yang, Qingyue Zhao, Quanquan Gu | Global Convergence and Rich Feature Learning in $L$-Layer Infinite-Width
Neural Networks under $\mu$P Parametrization | 29 pages, 5 figures, 2 tables | null | null | null | cs.LG cs.AI math.OC stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite deep neural networks' powerful representation learning capabilities,
theoretical understanding of how networks can simultaneously achieve meaningful
feature learning and global convergence remains elusive. Existing approaches
like the neural tangent kernel (NTK) are limited because features stay close to
their initialization in this parametrization, leaving open questions about
feature properties during substantial evolution. In this paper, we investigate
the training dynamics of infinitely wide, $L$-layer neural networks using the
tensor program (TP) framework. Specifically, we show that, when trained with
stochastic gradient descent (SGD) under the Maximal Update parametrization
($\mu$P) and mild conditions on the activation function, SGD enables these
networks to learn linearly independent features that substantially deviate from
their initial values. This rich feature space captures relevant data
information and ensures that any convergent point of the training process is a
global minimum. Our analysis leverages both the interactions among features
across layers and the properties of Gaussian random variables, providing new
insights into deep representation learning. We further validate our theoretical
findings through experiments on real-world datasets.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 17:33:13 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Chen",
"Zixiang",
""
],
[
"Yang",
"Greg",
""
],
[
"Zhao",
"Qingyue",
""
],
[
"Gu",
"Quanquan",
""
]
]
| TITLE: Global Convergence and Rich Feature Learning in $L$-Layer Infinite-Width
Neural Networks under $\mu$P Parametrization
ABSTRACT: Despite deep neural networks' powerful representation learning capabilities,
theoretical understanding of how networks can simultaneously achieve meaningful
feature learning and global convergence remains elusive. Existing approaches
like the neural tangent kernel (NTK) are limited because features stay close to
their initialization in this parametrization, leaving open questions about
feature properties during substantial evolution. In this paper, we investigate
the training dynamics of infinitely wide, $L$-layer neural networks using the
tensor program (TP) framework. Specifically, we show that, when trained with
stochastic gradient descent (SGD) under the Maximal Update parametrization
($\mu$P) and mild conditions on the activation function, SGD enables these
networks to learn linearly independent features that substantially deviate from
their initial values. This rich feature space captures relevant data
information and ensures that any convergent point of the training process is a
global minimum. Our analysis leverages both the interactions among features
across layers and the properties of Gaussian random variables, providing new
insights into deep representation learning. We further validate our theoretical
findings through experiments on real-world datasets.
| no_new_dataset | 0.945951 |
2503.09576 | Philippe Chlenski | Philippe Chlenski, Kaizhu Du, Dylan Satow, and Itsik Pe'er | Manify: A Python Library for Learning Non-Euclidean Representations | 30 pages, 4 figures, 4 tables. Preprint | null | null | null | cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | We present Manify, an open-source Python library for non-Euclidean
representation learning. Leveraging manifold learning techniques, Manify
provides tools for learning embeddings in (products of) non-Euclidean spaces,
performing classification and regression with data that lives in such spaces,
and estimating the curvature of a manifold. Manify aims to advance research and
applications in machine learning by offering a comprehensive suite of tools for
manifold-based data analysis. Our source code, examples, datasets, results, and
documentation are available at https://github.com/pchlenski/manify
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 17:44:40 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Chlenski",
"Philippe",
""
],
[
"Du",
"Kaizhu",
""
],
[
"Satow",
"Dylan",
""
],
[
"Pe'er",
"Itsik",
""
]
]
| TITLE: Manify: A Python Library for Learning Non-Euclidean Representations
ABSTRACT: We present Manify, an open-source Python library for non-Euclidean
representation learning. Leveraging manifold learning techniques, Manify
provides tools for learning embeddings in (products of) non-Euclidean spaces,
performing classification and regression with data that lives in such spaces,
and estimating the curvature of a manifold. Manify aims to advance research and
applications in machine learning by offering a comprehensive suite of tools for
manifold-based data analysis. Our source code, examples, datasets, results, and
documentation are available at https://github.com/pchlenski/manify
| no_new_dataset | 0.944177 |
2503.09587 | Nannan Wu | Nannan Wu, Zhuo Kuang, Zengqiang Yan, Ping Wang, Li Yu | Fair Federated Medical Image Classification Against Quality Shift via
Inter-Client Progressive State Matching | Preprint | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the potential of federated learning in medical applications,
inconsistent imaging quality across institutions-stemming from lower-quality
data from a minority of clients-biases federated models toward more common
high-quality images. This raises significant fairness concerns. Existing fair
federated learning methods have demonstrated some effectiveness in solving this
problem by aligning a single 0th- or 1st-order state of convergence (e.g.,
training loss or sharpness). However, we argue in this work that fairness based
on such a single state is still not an adequate surrogate for fairness during
testing, as these single metrics fail to fully capture the convergence
characteristics, making them suboptimal for guiding fair learning. To address
this limitation, we develop a generalized framework. Specifically, we propose
assessing convergence using multiple states, defined as sharpness or perturbed
loss computed at varying search distances. Building on this comprehensive
assessment, we propose promoting fairness for these states across clients to
achieve our ultimate fairness objective. This is accomplished through the
proposed method, FedISM+. In FedISM+, the search distance evolves over time,
progressively focusing on different states. We then incorporate two components
in local training and global aggregation to ensure cross-client fairness for
each state. This gradually makes convergence equitable for all states, thereby
improving fairness during testing. Our empirical evaluations, performed on the
well-known RSNA ICH and ISIC 2019 datasets, demonstrate the superiority of
FedISM+ over existing state-of-the-art methods for fair federated learning. The
code is available at https://github.com/wnn2000/FFL4MIA.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 17:56:28 GMT"
}
]
| 2025-03-13T00:00:00 | [
[
"Wu",
"Nannan",
""
],
[
"Kuang",
"Zhuo",
""
],
[
"Yan",
"Zengqiang",
""
],
[
"Wang",
"Ping",
""
],
[
"Yu",
"Li",
""
]
]
| TITLE: Fair Federated Medical Image Classification Against Quality Shift via
Inter-Client Progressive State Matching
ABSTRACT: Despite the potential of federated learning in medical applications,
inconsistent imaging quality across institutions-stemming from lower-quality
data from a minority of clients-biases federated models toward more common
high-quality images. This raises significant fairness concerns. Existing fair
federated learning methods have demonstrated some effectiveness in solving this
problem by aligning a single 0th- or 1st-order state of convergence (e.g.,
training loss or sharpness). However, we argue in this work that fairness based
on such a single state is still not an adequate surrogate for fairness during
testing, as these single metrics fail to fully capture the convergence
characteristics, making them suboptimal for guiding fair learning. To address
this limitation, we develop a generalized framework. Specifically, we propose
assessing convergence using multiple states, defined as sharpness or perturbed
loss computed at varying search distances. Building on this comprehensive
assessment, we propose promoting fairness for these states across clients to
achieve our ultimate fairness objective. This is accomplished through the
proposed method, FedISM+. In FedISM+, the search distance evolves over time,
progressively focusing on different states. We then incorporate two components
in local training and global aggregation to ensure cross-client fairness for
each state. This gradually makes convergence equitable for all states, thereby
improving fairness during testing. Our empirical evaluations, performed on the
well-known RSNA ICH and ISIC 2019 datasets, demonstrate the superiority of
FedISM+ over existing state-of-the-art methods for fair federated learning. The
code is available at https://github.com/wnn2000/FFL4MIA.
| no_new_dataset | 0.947332 |
2103.01901 | Shuxiao Chen | Shuxiao Chen, Qinqing Zheng, Qi Long, Weijie J. Su | Minimax Estimation for Personalized Federated Learning: An Alternative
between FedAvg and Local Training? | JMLR published version | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A widely recognized difficulty in federated learning arises from the
statistical heterogeneity among clients: local datasets often originate from
distinct yet not entirely unrelated probability distributions, and
personalization is, therefore, necessary to achieve optimal results from each
individual's perspective. In this paper, we show how the excess risks of
personalized federated learning using a smooth, strongly convex loss depend on
data heterogeneity from a minimax point of view, with a focus on the FedAvg
algorithm (McMahan et al., 2017) and pure local training (i.e., clients solve
empirical risk minimization problems on their local datasets without any
communication). Our main result reveals an approximate alternative between
these two baseline algorithms for federated learning: the former algorithm is
minimax rate optimal over a collection of instances when data heterogeneity is
small, whereas the latter is minimax rate optimal when data heterogeneity is
large, and the threshold is sharp up to a constant.
As an implication, our results show that from a worst-case point of view, a
dichotomous strategy that makes a choice between the two baseline algorithms is
rate-optimal. Another implication is that the popular FedAvg following by local
fine tuning strategy is also minimax optimal under additional regularity
conditions. Our analysis relies on a new notion of algorithmic stability that
takes into account the nature of federated learning.
| [
{
"version": "v1",
"created": "Tue, 2 Mar 2021 17:58:20 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 02:36:12 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Chen",
"Shuxiao",
""
],
[
"Zheng",
"Qinqing",
""
],
[
"Long",
"Qi",
""
],
[
"Su",
"Weijie J.",
""
]
]
| TITLE: Minimax Estimation for Personalized Federated Learning: An Alternative
between FedAvg and Local Training?
ABSTRACT: A widely recognized difficulty in federated learning arises from the
statistical heterogeneity among clients: local datasets often originate from
distinct yet not entirely unrelated probability distributions, and
personalization is, therefore, necessary to achieve optimal results from each
individual's perspective. In this paper, we show how the excess risks of
personalized federated learning using a smooth, strongly convex loss depend on
data heterogeneity from a minimax point of view, with a focus on the FedAvg
algorithm (McMahan et al., 2017) and pure local training (i.e., clients solve
empirical risk minimization problems on their local datasets without any
communication). Our main result reveals an approximate alternative between
these two baseline algorithms for federated learning: the former algorithm is
minimax rate optimal over a collection of instances when data heterogeneity is
small, whereas the latter is minimax rate optimal when data heterogeneity is
large, and the threshold is sharp up to a constant.
As an implication, our results show that from a worst-case point of view, a
dichotomous strategy that makes a choice between the two baseline algorithms is
rate-optimal. Another implication is that the popular FedAvg following by local
fine tuning strategy is also minimax optimal under additional regularity
conditions. Our analysis relies on a new notion of algorithmic stability that
takes into account the nature of federated learning.
| no_new_dataset | 0.950595 |
2207.05195 | Bohan Tang | Bohan Tang, Yiqi Zhong, Chenxin Xu, Wei-Tao Wu, Ulrich Neumann,
Yanfeng Wang, Ya Zhang, and Siheng Chen | Collaborative Uncertainty Benefits Multi-Agent Multi-Modal Trajectory
Forecasting | arXiv admin note: text overlap with arXiv:2110.13947 | null | null | null | cs.CV stat.ML | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In multi-modal multi-agent trajectory forecasting, two major challenges have
not been fully tackled: 1) how to measure the uncertainty brought by the
interaction module that causes correlations among the predicted trajectories of
multiple agents; 2) how to rank the multiple predictions and select the optimal
predicted trajectory. In order to handle these challenges, this work first
proposes a novel concept, collaborative uncertainty (CU), which models the
uncertainty resulting from interaction modules. Then we build a general
CU-aware regression framework with an original permutation-equivariant
uncertainty estimator to do both tasks of regression and uncertainty
estimation. Further, we apply the proposed framework to current SOTA
multi-agent multi-modal forecasting systems as a plugin module, which enables
the SOTA systems to 1) estimate the uncertainty in the multi-agent multi-modal
trajectory forecasting task; 2) rank the multiple predictions and select the
optimal one based on the estimated uncertainty. We conduct extensive
experiments on a synthetic dataset and two public large-scale multi-agent
trajectory forecasting benchmarks. Experiments show that: 1) on the synthetic
dataset, the CU-aware regression framework allows the model to appropriately
approximate the ground-truth Laplace distribution; 2) on the multi-agent
trajectory forecasting benchmarks, the CU-aware regression framework steadily
helps SOTA systems improve their performances. Specially, the proposed
framework helps VectorNet improve by 262 cm regarding the Final Displacement
Error of the chosen optimal prediction on the nuScenes dataset; 3) for
multi-agent multi-modal trajectory forecasting systems, prediction uncertainty
is positively correlated with future stochasticity; and 4) the estimated CU
values are highly related to the interactive information among agents.
| [
{
"version": "v1",
"created": "Mon, 11 Jul 2022 21:17:41 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 16:10:29 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Tang",
"Bohan",
""
],
[
"Zhong",
"Yiqi",
""
],
[
"Xu",
"Chenxin",
""
],
[
"Wu",
"Wei-Tao",
""
],
[
"Neumann",
"Ulrich",
""
],
[
"Wang",
"Yanfeng",
""
],
[
"Zhang",
"Ya",
""
],
[
"Chen",
"Siheng",
""
]
]
| TITLE: Collaborative Uncertainty Benefits Multi-Agent Multi-Modal Trajectory
Forecasting
ABSTRACT: In multi-modal multi-agent trajectory forecasting, two major challenges have
not been fully tackled: 1) how to measure the uncertainty brought by the
interaction module that causes correlations among the predicted trajectories of
multiple agents; 2) how to rank the multiple predictions and select the optimal
predicted trajectory. In order to handle these challenges, this work first
proposes a novel concept, collaborative uncertainty (CU), which models the
uncertainty resulting from interaction modules. Then we build a general
CU-aware regression framework with an original permutation-equivariant
uncertainty estimator to do both tasks of regression and uncertainty
estimation. Further, we apply the proposed framework to current SOTA
multi-agent multi-modal forecasting systems as a plugin module, which enables
the SOTA systems to 1) estimate the uncertainty in the multi-agent multi-modal
trajectory forecasting task; 2) rank the multiple predictions and select the
optimal one based on the estimated uncertainty. We conduct extensive
experiments on a synthetic dataset and two public large-scale multi-agent
trajectory forecasting benchmarks. Experiments show that: 1) on the synthetic
dataset, the CU-aware regression framework allows the model to appropriately
approximate the ground-truth Laplace distribution; 2) on the multi-agent
trajectory forecasting benchmarks, the CU-aware regression framework steadily
helps SOTA systems improve their performances. Specially, the proposed
framework helps VectorNet improve by 262 cm regarding the Final Displacement
Error of the chosen optimal prediction on the nuScenes dataset; 3) for
multi-agent multi-modal trajectory forecasting systems, prediction uncertainty
is positively correlated with future stochasticity; and 4) the estimated CU
values are highly related to the interactive information among agents.
| no_new_dataset | 0.946941 |
2211.01717 | Bohan Tang | Bohan Tang, Siheng Chen, Xiaowen Dong | Learning Hypergraphs From Signals With Dual Smoothness Prior | null | null | null | null | cs.LG cs.SI eess.SP stat.ML | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Hypergraph structure learning, which aims to learn the hypergraph structures
from the observed signals to capture the intrinsic high-order relationships
among the entities, becomes crucial when a hypergraph topology is not readily
available in the datasets. There are two challenges that lie at the heart of
this problem: 1) how to handle the huge search space of potential hyperedges,
and 2) how to define meaningful criteria to measure the relationship between
the signals observed on nodes and the hypergraph structure. In this paper, for
the first challenge, we adopt the assumption that the ideal hypergraph
structure can be derived from a learnable graph structure that captures the
pairwise relations within signals. Further, we propose a hypergraph structure
learning framework HGSL with a novel dual smoothness prior that reveals a
mapping between the observed node signals and the hypergraph structure, whereby
each hyperedge corresponds to a subgraph with both node signal smoothness and
edge signal smoothness in the learnable graph structure. Finally, we conduct
extensive experiments to evaluate HGSL on both synthetic and real world
datasets. Experiments show that HGSL can efficiently infer meaningful
hypergraph topologies from observed signals.
| [
{
"version": "v1",
"created": "Thu, 3 Nov 2022 11:13:02 GMT"
},
{
"version": "v2",
"created": "Thu, 16 Feb 2023 10:44:48 GMT"
},
{
"version": "v3",
"created": "Tue, 14 Mar 2023 10:21:56 GMT"
},
{
"version": "v4",
"created": "Tue, 11 Mar 2025 16:08:58 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Tang",
"Bohan",
""
],
[
"Chen",
"Siheng",
""
],
[
"Dong",
"Xiaowen",
""
]
]
| TITLE: Learning Hypergraphs From Signals With Dual Smoothness Prior
ABSTRACT: Hypergraph structure learning, which aims to learn the hypergraph structures
from the observed signals to capture the intrinsic high-order relationships
among the entities, becomes crucial when a hypergraph topology is not readily
available in the datasets. There are two challenges that lie at the heart of
this problem: 1) how to handle the huge search space of potential hyperedges,
and 2) how to define meaningful criteria to measure the relationship between
the signals observed on nodes and the hypergraph structure. In this paper, for
the first challenge, we adopt the assumption that the ideal hypergraph
structure can be derived from a learnable graph structure that captures the
pairwise relations within signals. Further, we propose a hypergraph structure
learning framework HGSL with a novel dual smoothness prior that reveals a
mapping between the observed node signals and the hypergraph structure, whereby
each hyperedge corresponds to a subgraph with both node signal smoothness and
edge signal smoothness in the learnable graph structure. Finally, we conduct
extensive experiments to evaluate HGSL on both synthetic and real world
datasets. Experiments show that HGSL can efficiently infer meaningful
hypergraph topologies from observed signals.
| no_new_dataset | 0.949763 |
2301.08995 | Anoop Kadan | Anoop Kadan, Deepak P., Manjary P. Gangan, Savitha Sam Abraham, Lajish
V. L | REDAffectiveLM: Leveraging Affect Enriched Embedding and
Transformer-based Neural Language Model for Readers' Emotion Detection | null | null | 10.1007/s10115-024-02194-4 | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Technological advancements in web platforms allow people to express and share
emotions towards textual write-ups written and shared by others. This brings
about different interesting domains for analysis; emotion expressed by the
writer and emotion elicited from the readers. In this paper, we propose a novel
approach for Readers' Emotion Detection from short-text documents using a deep
learning model called REDAffectiveLM. Within state-of-the-art NLP tasks, it is
well understood that utilizing context-specific representations from
transformer-based pre-trained language models helps achieve improved
performance. Within this affective computing task, we explore how incorporating
affective information can further enhance performance. Towards this, we
leverage context-specific and affect enriched representations by using a
transformer-based pre-trained language model in tandem with affect enriched
Bi-LSTM+Attention. For empirical evaluation, we procure a new dataset REN-20k,
besides using RENh-4k and SemEval-2007. We evaluate the performance of our
REDAffectiveLM rigorously across these datasets, against a vast set of
state-of-the-art baselines, where our model consistently outperforms baselines
and obtains statistically significant results. Our results establish that
utilizing affect enriched representation along with context-specific
representation within a neural architecture can considerably enhance readers'
emotion detection. Since the impact of affect enrichment specifically in
readers' emotion detection isn't well explored, we conduct a detailed analysis
over affect enriched Bi-LSTM+Attention using qualitative and quantitative model
behavior evaluation techniques. We observe that compared to conventional
semantic embedding, affect enriched embedding increases ability of the network
to effectively identify and assign weightage to key terms responsible for
readers' emotion detection.
| [
{
"version": "v1",
"created": "Sat, 21 Jan 2023 19:28:25 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Kadan",
"Anoop",
""
],
[
"P.",
"Deepak",
""
],
[
"Gangan",
"Manjary P.",
""
],
[
"Abraham",
"Savitha Sam",
""
],
[
"L",
"Lajish V.",
""
]
]
| TITLE: REDAffectiveLM: Leveraging Affect Enriched Embedding and
Transformer-based Neural Language Model for Readers' Emotion Detection
ABSTRACT: Technological advancements in web platforms allow people to express and share
emotions towards textual write-ups written and shared by others. This brings
about different interesting domains for analysis; emotion expressed by the
writer and emotion elicited from the readers. In this paper, we propose a novel
approach for Readers' Emotion Detection from short-text documents using a deep
learning model called REDAffectiveLM. Within state-of-the-art NLP tasks, it is
well understood that utilizing context-specific representations from
transformer-based pre-trained language models helps achieve improved
performance. Within this affective computing task, we explore how incorporating
affective information can further enhance performance. Towards this, we
leverage context-specific and affect enriched representations by using a
transformer-based pre-trained language model in tandem with affect enriched
Bi-LSTM+Attention. For empirical evaluation, we procure a new dataset REN-20k,
besides using RENh-4k and SemEval-2007. We evaluate the performance of our
REDAffectiveLM rigorously across these datasets, against a vast set of
state-of-the-art baselines, where our model consistently outperforms baselines
and obtains statistically significant results. Our results establish that
utilizing affect enriched representation along with context-specific
representation within a neural architecture can considerably enhance readers'
emotion detection. Since the impact of affect enrichment specifically in
readers' emotion detection isn't well explored, we conduct a detailed analysis
over affect enriched Bi-LSTM+Attention using qualitative and quantitative model
behavior evaluation techniques. We observe that compared to conventional
semantic embedding, affect enriched embedding increases ability of the network
to effectively identify and assign weightage to key terms responsible for
readers' emotion detection.
| new_dataset | 0.966632 |
2305.18226 | Christoforos Vasilatos | Christoforos Vasilatos, Manaar Alam, Talal Rahwan, Yasir Zaki and
Michail Maniatakos | HowkGPT: Investigating the Detection of ChatGPT-generated University
Student Homework through Context-Aware Perplexity Analysis | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As the use of Large Language Models (LLMs) in text generation tasks
proliferates, concerns arise over their potential to compromise academic
integrity. The education sector currently tussles with distinguishing
student-authored homework assignments from AI-generated ones. This paper
addresses the challenge by introducing HowkGPT, designed to identify homework
assignments generated by AI. HowkGPT is built upon a dataset of academic
assignments and accompanying metadata [17] and employs a pretrained LLM to
compute perplexity scores for student-authored and ChatGPT-generated responses.
These scores then assist in establishing a threshold for discerning the origin
of a submitted assignment. Given the specificity and contextual nature of
academic work, HowkGPT further refines its analysis by defining
category-specific thresholds derived from the metadata, enhancing the precision
of the detection. This study emphasizes the critical need for effective
strategies to uphold academic integrity amidst the growing influence of LLMs
and provides an approach to ensuring fair and accurate grading in educational
institutions.
| [
{
"version": "v1",
"created": "Fri, 26 May 2023 11:07:25 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 11:43:44 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 08:08:05 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Vasilatos",
"Christoforos",
""
],
[
"Alam",
"Manaar",
""
],
[
"Rahwan",
"Talal",
""
],
[
"Zaki",
"Yasir",
""
],
[
"Maniatakos",
"Michail",
""
]
]
| TITLE: HowkGPT: Investigating the Detection of ChatGPT-generated University
Student Homework through Context-Aware Perplexity Analysis
ABSTRACT: As the use of Large Language Models (LLMs) in text generation tasks
proliferates, concerns arise over their potential to compromise academic
integrity. The education sector currently tussles with distinguishing
student-authored homework assignments from AI-generated ones. This paper
addresses the challenge by introducing HowkGPT, designed to identify homework
assignments generated by AI. HowkGPT is built upon a dataset of academic
assignments and accompanying metadata [17] and employs a pretrained LLM to
compute perplexity scores for student-authored and ChatGPT-generated responses.
These scores then assist in establishing a threshold for discerning the origin
of a submitted assignment. Given the specificity and contextual nature of
academic work, HowkGPT further refines its analysis by defining
category-specific thresholds derived from the metadata, enhancing the precision
of the detection. This study emphasizes the critical need for effective
strategies to uphold academic integrity amidst the growing influence of LLMs
and provides an approach to ensuring fair and accurate grading in educational
institutions.
| new_dataset | 0.96738 |
2306.01740 | John Cartlidge | Lawrence Clegg and John Cartlidge | Not feeling the buzz: Correction study of mispricing and inefficiency in
online sportsbooks | 24 pages, 2 figures. Revised argument; minor edits and funding
acknowledgement; typos corrected. This paper is a replication and correction
study and longer version of an accepted journal article. Replication code and
data are available online: https://github.com/Faxulous/notFeelingTheBuzz | International Journal of Forecasting (2025), 41(2), pages 798-802 | 10.1016/j.ijforecast.2024.06.012 | null | stat.AP cs.CE q-fin.GN q-fin.ST | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a replication and correction of a recent article (Ramirez, P.,
Reade, J.J., Singleton, C., Betting on a buzz: Mispricing and inefficiency in
online sportsbooks, International Journal of Forecasting, 39:3, 2023, pp.
1413-1423, doi: 10.1016/j.ijforecast.2022.07.011). RRS measure profile page
views on Wikipedia to generate a "buzz factor" metric for tennis players and
show that it can be used to form a profitable gambling strategy by predicting
bookmaker mispricing. Here, we use the same dataset as RRS to reproduce their
results exactly, thus confirming the robustness of their mispricing claim.
However, we discover that the published betting results are significantly
affected by a single bet (the "Hercog" bet), which returns substantial outlier
profits based on erroneously long odds. When this data quality issue is
resolved, the majority of reported profits disappear and only one strategy,
which bets on "competitive" matches, remains significantly profitable in the
original out-of-sample period. While one profitable strategy offers weaker
support than the original study, it still provides an indication that market
inefficiencies may exist, as originally claimed by RRS. As an extension, we
continue backtesting after 2020 on a cleaned dataset. Results show that (a) the
"competitive" strategy generates no further profits, potentially suggesting
markets have become more efficient, and (b) model coefficients estimated over
this more recent period are no longer reliable predictors of bookmaker
mispricing. We present this work as a case study demonstrating the importance
of replication studies in sports forecasting, and the necessity to clean data.
We open-source release comprehensive datasets and code.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 19:25:33 GMT"
},
{
"version": "v2",
"created": "Thu, 25 Jan 2024 16:33:33 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jun 2024 09:03:30 GMT"
},
{
"version": "v4",
"created": "Thu, 11 Jul 2024 14:50:38 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Clegg",
"Lawrence",
""
],
[
"Cartlidge",
"John",
""
]
]
| TITLE: Not feeling the buzz: Correction study of mispricing and inefficiency in
online sportsbooks
ABSTRACT: We present a replication and correction of a recent article (Ramirez, P.,
Reade, J.J., Singleton, C., Betting on a buzz: Mispricing and inefficiency in
online sportsbooks, International Journal of Forecasting, 39:3, 2023, pp.
1413-1423, doi: 10.1016/j.ijforecast.2022.07.011). RRS measure profile page
views on Wikipedia to generate a "buzz factor" metric for tennis players and
show that it can be used to form a profitable gambling strategy by predicting
bookmaker mispricing. Here, we use the same dataset as RRS to reproduce their
results exactly, thus confirming the robustness of their mispricing claim.
However, we discover that the published betting results are significantly
affected by a single bet (the "Hercog" bet), which returns substantial outlier
profits based on erroneously long odds. When this data quality issue is
resolved, the majority of reported profits disappear and only one strategy,
which bets on "competitive" matches, remains significantly profitable in the
original out-of-sample period. While one profitable strategy offers weaker
support than the original study, it still provides an indication that market
inefficiencies may exist, as originally claimed by RRS. As an extension, we
continue backtesting after 2020 on a cleaned dataset. Results show that (a) the
"competitive" strategy generates no further profits, potentially suggesting
markets have become more efficient, and (b) model coefficients estimated over
this more recent period are no longer reliable predictors of bookmaker
mispricing. We present this work as a case study demonstrating the importance
of replication studies in sports forecasting, and the necessity to clean data.
We open-source release comprehensive datasets and code.
| no_new_dataset | 0.922552 |
2307.08359 | Frederik Plahl | Andreas Zachariae and Julia Widera and Frederik Plahl and Bj\"orn Hein
and Christian Wurll | Human Emergency Detection during Autonomous Hospital Transports | Preprint of the corresponding IAS18-2023 conference publication
(Proceedings of the 18th International Conference IAS-18) | Lee, SG., An, J., Chong, N.Y., Strand, M., Kim, J.H. (eds)
Intelligent Autonomous Systems 18. IAS 2023. Lecture Notes in Networks and
Systems, vol 794. Springer, Cham | 10.1007/978-3-031-44981-9_21 | null | cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Human transports in hospitals are labor-intensive and primarily performed in
beds to save time. This transfer method does not promote the mobility or
autonomy of the patient. To relieve the caregivers from this time-consuming
task, a mobile robot is developed to autonomously transport humans around the
hospital. It provides different transfer modes including walking and sitting in
a wheelchair. The problem that this paper focuses on is to detect emergencies
and ensure the well-being of the patient during the transport. For this
purpose, the patient is tracked and monitored with a camera system. OpenPose is
used for Human Pose Estimation and a trained classifier for emergency
detection. We collected and published a dataset of 18,000 images in lab and
hospital environments. It differs from related work because we have a moving
robot with different transfer modes in a highly dynamic environment with
multiple people in the scene using only RGB-D data. To improve the critical
recall metric, we apply threshold moving and a time delay. We compare different
models with an AutoML approach. This paper shows that emergencies while walking
are best detected by a SVM with a recall of 95.8% on single frames. In the case
of sitting transport, the best model achieves a recall of 62.2%. The
contribution is to establish a baseline on this new dataset and to provide a
proof of concept for the human emergency detection in this use case.
| [
{
"version": "v1",
"created": "Mon, 17 Jul 2023 09:54:52 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Zachariae",
"Andreas",
""
],
[
"Widera",
"Julia",
""
],
[
"Plahl",
"Frederik",
""
],
[
"Hein",
"Björn",
""
],
[
"Wurll",
"Christian",
""
]
]
| TITLE: Human Emergency Detection during Autonomous Hospital Transports
ABSTRACT: Human transports in hospitals are labor-intensive and primarily performed in
beds to save time. This transfer method does not promote the mobility or
autonomy of the patient. To relieve the caregivers from this time-consuming
task, a mobile robot is developed to autonomously transport humans around the
hospital. It provides different transfer modes including walking and sitting in
a wheelchair. The problem that this paper focuses on is to detect emergencies
and ensure the well-being of the patient during the transport. For this
purpose, the patient is tracked and monitored with a camera system. OpenPose is
used for Human Pose Estimation and a trained classifier for emergency
detection. We collected and published a dataset of 18,000 images in lab and
hospital environments. It differs from related work because we have a moving
robot with different transfer modes in a highly dynamic environment with
multiple people in the scene using only RGB-D data. To improve the critical
recall metric, we apply threshold moving and a time delay. We compare different
models with an AutoML approach. This paper shows that emergencies while walking
are best detected by a SVM with a recall of 95.8% on single frames. In the case
of sitting transport, the best model achieves a recall of 62.2%. The
contribution is to establish a baseline on this new dataset and to provide a
proof of concept for the human emergency detection in this use case.
| new_dataset | 0.958499 |
2308.11912 | Soonwoo Kwon | Soonwoo Kwon, Sojung Kim, Seunghyun Lee, Jin-Young Kim, Suyeong An,
and Kyuseok Kim | Addressing Selection Bias in Computerized Adaptive Testing: A User-Wise
Aggregate Influence Function Approach | CIKM 2023 | null | null | null | cs.LG cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Computerized Adaptive Testing (CAT) is a widely used, efficient test mode
that adapts to the examinee's proficiency level in the test domain. CAT
requires pre-trained item profiles, for CAT iteratively assesses the student
real-time based on the registered items' profiles, and selects the next item to
administer using candidate items' profiles. However, obtaining such item
profiles is a costly process that involves gathering a large, dense
item-response data, then training a diagnostic model on the collected data. In
this paper, we explore the possibility of leveraging response data collected in
the CAT service. We first show that this poses a unique challenge due to the
inherent selection bias introduced by CAT, i.e., more proficient students will
receive harder questions. Indeed, when naively training the diagnostic model
using CAT response data, we observe that item profiles deviate significantly
from the ground-truth. To tackle the selection bias issue, we propose the
user-wise aggregate influence function method. Our intuition is to filter out
users whose response data is heavily biased in an aggregate manner, as judged
by how much perturbation the added data will introduce during parameter
estimation. This way, we may enhance the performance of CAT while introducing
minimal bias to the item profiles. We provide extensive experiments to
demonstrate the superiority of our proposed method based on the three public
datasets and one dataset that contains real-world CAT response data.
| [
{
"version": "v1",
"created": "Wed, 23 Aug 2023 04:57:21 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Kwon",
"Soonwoo",
""
],
[
"Kim",
"Sojung",
""
],
[
"Lee",
"Seunghyun",
""
],
[
"Kim",
"Jin-Young",
""
],
[
"An",
"Suyeong",
""
],
[
"Kim",
"Kyuseok",
""
]
]
| TITLE: Addressing Selection Bias in Computerized Adaptive Testing: A User-Wise
Aggregate Influence Function Approach
ABSTRACT: Computerized Adaptive Testing (CAT) is a widely used, efficient test mode
that adapts to the examinee's proficiency level in the test domain. CAT
requires pre-trained item profiles, for CAT iteratively assesses the student
real-time based on the registered items' profiles, and selects the next item to
administer using candidate items' profiles. However, obtaining such item
profiles is a costly process that involves gathering a large, dense
item-response data, then training a diagnostic model on the collected data. In
this paper, we explore the possibility of leveraging response data collected in
the CAT service. We first show that this poses a unique challenge due to the
inherent selection bias introduced by CAT, i.e., more proficient students will
receive harder questions. Indeed, when naively training the diagnostic model
using CAT response data, we observe that item profiles deviate significantly
from the ground-truth. To tackle the selection bias issue, we propose the
user-wise aggregate influence function method. Our intuition is to filter out
users whose response data is heavily biased in an aggregate manner, as judged
by how much perturbation the added data will introduce during parameter
estimation. This way, we may enhance the performance of CAT while introducing
minimal bias to the item profiles. We provide extensive experiments to
demonstrate the superiority of our proposed method based on the three public
datasets and one dataset that contains real-world CAT response data.
| no_new_dataset | 0.913599 |
2309.12862 | Yuwei Sun | Yuwei Sun, Hideya Ochiai, Zhirong Wu, Stephen Lin, Ryota Kanai | Associative Transformer | Accepted for CVPR 2025 | null | null | null | cs.LG cs.CV cs.NE | http://creativecommons.org/licenses/by/4.0/ | Emerging from the pairwise attention in conventional Transformers, there is a
growing interest in sparse attention mechanisms that align more closely with
localized, contextual learning in the biological brain. Existing studies such
as the Coordination method employ iterative cross-attention mechanisms with a
bottleneck to enable the sparse association of inputs. However, these methods
are parameter inefficient and fail in more complex relational reasoning tasks.
To this end, we propose Associative Transformer (AiT) to enhance the
association among sparsely attended input tokens, improving parameter
efficiency and performance in various vision tasks such as classification and
relational reasoning. AiT leverages a learnable explicit memory comprising
specialized priors that guide bottleneck attentions to facilitate the
extraction of diverse localized tokens. Moreover, AiT employs an associative
memory-based token reconstruction using a Hopfield energy function. The
extensive empirical experiments demonstrate that AiT requires significantly
fewer parameters and attention layers outperforming a broad range of sparse
Transformer models. Additionally, AiT outperforms the SOTA sparse Transformer
models including the Coordination method on the Sort-of-CLEVR dataset.
| [
{
"version": "v1",
"created": "Fri, 22 Sep 2023 13:37:10 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Nov 2023 07:26:55 GMT"
},
{
"version": "v3",
"created": "Wed, 31 Jan 2024 01:05:14 GMT"
},
{
"version": "v4",
"created": "Tue, 11 Mar 2025 09:04:22 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Sun",
"Yuwei",
""
],
[
"Ochiai",
"Hideya",
""
],
[
"Wu",
"Zhirong",
""
],
[
"Lin",
"Stephen",
""
],
[
"Kanai",
"Ryota",
""
]
]
| TITLE: Associative Transformer
ABSTRACT: Emerging from the pairwise attention in conventional Transformers, there is a
growing interest in sparse attention mechanisms that align more closely with
localized, contextual learning in the biological brain. Existing studies such
as the Coordination method employ iterative cross-attention mechanisms with a
bottleneck to enable the sparse association of inputs. However, these methods
are parameter inefficient and fail in more complex relational reasoning tasks.
To this end, we propose Associative Transformer (AiT) to enhance the
association among sparsely attended input tokens, improving parameter
efficiency and performance in various vision tasks such as classification and
relational reasoning. AiT leverages a learnable explicit memory comprising
specialized priors that guide bottleneck attentions to facilitate the
extraction of diverse localized tokens. Moreover, AiT employs an associative
memory-based token reconstruction using a Hopfield energy function. The
extensive empirical experiments demonstrate that AiT requires significantly
fewer parameters and attention layers outperforming a broad range of sparse
Transformer models. Additionally, AiT outperforms the SOTA sparse Transformer
models including the Coordination method on the Sort-of-CLEVR dataset.
| no_new_dataset | 0.945701 |
2310.03617 | Yihong Tang | Yihong Tang, Zhan Zhao, Weipeng Deng, Shuyu Lei, Yuebing Liang,
Zhenliang Ma | RouteKG: A knowledge graph-based framework for route prediction on road
networks | null | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Short-term route prediction on road networks allows us to anticipate the
future trajectories of road users, enabling various applications ranging from
dynamic traffic control to personalized navigation. Despite recent advances in
this area, existing methods focus primarily on learning sequential transition
patterns, neglecting the inherent spatial relations in road networks that can
affect human routing decisions. To fill this gap, this paper introduces
RouteKG, a novel Knowledge Graph-based framework for route prediction.
Specifically, we construct a Knowledge Graph on the road network to encode
spatial relations, especially moving directions that are crucial for human
navigation. Moreover, an n-ary tree-based algorithm is introduced to
efficiently generate top-K routes in a batch mode, enhancing computational
efficiency. To further optimize the prediction performance, a rank refinement
module is incorporated to fine-tune the candidate route rankings. The model
performance is evaluated using two real-world vehicle trajectory datasets from
two Chinese cities under various practical scenarios. The results demonstrate a
significant improvement in accuracy over baseline methods. We further validate
the proposed method by utilizing the pre-trained model as a simulator for
real-time traffic flow estimation at the link level. RouteKG holds great
potential for transforming vehicle navigation, traffic management, and a
variety of intelligent transportation tasks, playing a crucial role in
advancing the core foundation of intelligent and connected urban systems.
| [
{
"version": "v1",
"created": "Tue, 3 Oct 2023 10:40:35 GMT"
},
{
"version": "v2",
"created": "Sat, 24 Feb 2024 20:03:32 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Mar 2025 21:11:50 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Tang",
"Yihong",
""
],
[
"Zhao",
"Zhan",
""
],
[
"Deng",
"Weipeng",
""
],
[
"Lei",
"Shuyu",
""
],
[
"Liang",
"Yuebing",
""
],
[
"Ma",
"Zhenliang",
""
]
]
| TITLE: RouteKG: A knowledge graph-based framework for route prediction on road
networks
ABSTRACT: Short-term route prediction on road networks allows us to anticipate the
future trajectories of road users, enabling various applications ranging from
dynamic traffic control to personalized navigation. Despite recent advances in
this area, existing methods focus primarily on learning sequential transition
patterns, neglecting the inherent spatial relations in road networks that can
affect human routing decisions. To fill this gap, this paper introduces
RouteKG, a novel Knowledge Graph-based framework for route prediction.
Specifically, we construct a Knowledge Graph on the road network to encode
spatial relations, especially moving directions that are crucial for human
navigation. Moreover, an n-ary tree-based algorithm is introduced to
efficiently generate top-K routes in a batch mode, enhancing computational
efficiency. To further optimize the prediction performance, a rank refinement
module is incorporated to fine-tune the candidate route rankings. The model
performance is evaluated using two real-world vehicle trajectory datasets from
two Chinese cities under various practical scenarios. The results demonstrate a
significant improvement in accuracy over baseline methods. We further validate
the proposed method by utilizing the pre-trained model as a simulator for
real-time traffic flow estimation at the link level. RouteKG holds great
potential for transforming vehicle navigation, traffic management, and a
variety of intelligent transportation tasks, playing a crucial role in
advancing the core foundation of intelligent and connected urban systems.
| no_new_dataset | 0.942348 |
2310.10982 | Sepehr Samavi | Sepehr Samavi, James R. Han, Florian Shkurti, Angela P. Schoellig | SICNav: Safe and Interactive Crowd Navigation using Model Predictive
Control and Bilevel Optimization | Published in the IEEE Transactions on Robotics (T-RO) | Published 22 October 2024 | 10.1109/TRO.2024.3484634 | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robots need to predict and react to human motions to navigate through a crowd
without collisions. Many existing methods decouple prediction from planning,
which does not account for the interaction between robot and human motions and
can lead to the robot getting stuck. We propose SICNav, a Model Predictive
Control (MPC) method that jointly solves for robot motion and predicted crowd
motion in closed-loop. We model each human in the crowd to be following an
Optimal Reciprocal Collision Avoidance (ORCA) scheme and embed that model as a
constraint in the robot's local planner, resulting in a bilevel nonlinear MPC
optimization problem. We use a KKT-reformulation to cast the bilevel problem as
a single level and use a nonlinear solver to optimize. Our MPC method can
influence pedestrian motion while explicitly satisfying safety constraints in a
single-robot multi-human environment. We analyze the performance of SICNav in
two simulation environments and indoor experiments with a real robot to
demonstrate safe robot motion that can influence the surrounding humans. We
also validate the trajectory forecasting performance of ORCA on a human
trajectory dataset.
| [
{
"version": "v1",
"created": "Tue, 17 Oct 2023 04:07:08 GMT"
},
{
"version": "v2",
"created": "Mon, 27 May 2024 22:06:39 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 10:09:54 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Samavi",
"Sepehr",
""
],
[
"Han",
"James R.",
""
],
[
"Shkurti",
"Florian",
""
],
[
"Schoellig",
"Angela P.",
""
]
]
| TITLE: SICNav: Safe and Interactive Crowd Navigation using Model Predictive
Control and Bilevel Optimization
ABSTRACT: Robots need to predict and react to human motions to navigate through a crowd
without collisions. Many existing methods decouple prediction from planning,
which does not account for the interaction between robot and human motions and
can lead to the robot getting stuck. We propose SICNav, a Model Predictive
Control (MPC) method that jointly solves for robot motion and predicted crowd
motion in closed-loop. We model each human in the crowd to be following an
Optimal Reciprocal Collision Avoidance (ORCA) scheme and embed that model as a
constraint in the robot's local planner, resulting in a bilevel nonlinear MPC
optimization problem. We use a KKT-reformulation to cast the bilevel problem as
a single level and use a nonlinear solver to optimize. Our MPC method can
influence pedestrian motion while explicitly satisfying safety constraints in a
single-robot multi-human environment. We analyze the performance of SICNav in
two simulation environments and indoor experiments with a real robot to
demonstrate safe robot motion that can influence the surrounding humans. We
also validate the trajectory forecasting performance of ORCA on a human
trajectory dataset.
| no_new_dataset | 0.942771 |
2311.09350 | Wei-Di Chang | Wei-Di Chang, Francois Hogan, Scott Fujimoto, David Meger, and Gregory
Dudek | Generalizable Imitation Learning Through Pre-Trained Representations | ICRA 2025 Version | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | In this paper, we leverage self-supervised vision transformer models and
their emergent semantic abilities to improve the generalization abilities of
imitation learning policies. We introduce DVK, an imitation learning algorithm
that leverages rich pre-trained Visual Transformer patch-level embeddings to
obtain better generalization when learning through demonstrations. Our learner
sees the world by clustering appearance features into groups associated with
semantic concepts, forming stable keypoints that generalize across a wide range
of appearance variations and object types. We demonstrate how this
representation enables generalized behaviour by evaluating imitation learning
across a diverse dataset of object manipulation tasks. To facilitate further
study of generalization in Imitation Learning, all of our code for the method
and evaluation, as well as the dataset, is made available.
| [
{
"version": "v1",
"created": "Wed, 15 Nov 2023 20:15:51 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 18:57:28 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Chang",
"Wei-Di",
""
],
[
"Hogan",
"Francois",
""
],
[
"Fujimoto",
"Scott",
""
],
[
"Meger",
"David",
""
],
[
"Dudek",
"Gregory",
""
]
]
| TITLE: Generalizable Imitation Learning Through Pre-Trained Representations
ABSTRACT: In this paper, we leverage self-supervised vision transformer models and
their emergent semantic abilities to improve the generalization abilities of
imitation learning policies. We introduce DVK, an imitation learning algorithm
that leverages rich pre-trained Visual Transformer patch-level embeddings to
obtain better generalization when learning through demonstrations. Our learner
sees the world by clustering appearance features into groups associated with
semantic concepts, forming stable keypoints that generalize across a wide range
of appearance variations and object types. We demonstrate how this
representation enables generalized behaviour by evaluating imitation learning
across a diverse dataset of object manipulation tasks. To facilitate further
study of generalization in Imitation Learning, all of our code for the method
and evaluation, as well as the dataset, is made available.
| no_new_dataset | 0.940953 |
2311.10248 | Kan Yang | Sheldon C. Ebron, Meiying Zhang and Kan Yang | Identifying the Truth of Global Model: A Generic Solution to Defend
Against Byzantine and Backdoor Attacks in Federated Learning (full version) | Accepted to ACISP 2025. This is the full version | null | null | null | cs.LG cs.AI cs.CR cs.DC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Federated Learning (FL) enables multiple parties to train machine learning
models collaboratively without sharing the raw training data. However, the
federated nature of FL enables malicious clients to influence a trained model
by injecting error model updates via Byzantine or backdoor attacks. To detect
malicious model updates, a typical approach is to measure the distance between
each model update and a \textit{ground-truth model update}. To find such
\textit{ground-truth model updates}, existing defenses either require a benign
root dataset on the server (e.g., FLTrust) or simply use trimmed mean or median
as the threshold for clipping (e.g., FLAME). However, such benign root datasets
are impractical, and the trimmed mean or median may also eliminate
contributions from these underrepresented datasets.
In this paper, we propose a generic solution, namely FedTruth, to defend
against model poisoning attacks in FL, where the \textit{ground-truth model
update} (i.e., the global model update) will be estimated among all the model
updates with dynamic aggregation weights. Specifically, FedTruth does not have
specific assumptions on the benign or malicious data distribution or access to
a benign root dataset. Moreover, FedTruth considers the potential contributions
from all benign clients. Our empirical results show that FedTruth can reduce
the impacts of poisoned model updates against both Byzantine and backdoor
attacks, and is also efficient in large-scale FL systems.
| [
{
"version": "v1",
"created": "Fri, 17 Nov 2023 00:39:59 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 18:37:26 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Ebron",
"Sheldon C.",
""
],
[
"Zhang",
"Meiying",
""
],
[
"Yang",
"Kan",
""
]
]
| TITLE: Identifying the Truth of Global Model: A Generic Solution to Defend
Against Byzantine and Backdoor Attacks in Federated Learning (full version)
ABSTRACT: Federated Learning (FL) enables multiple parties to train machine learning
models collaboratively without sharing the raw training data. However, the
federated nature of FL enables malicious clients to influence a trained model
by injecting error model updates via Byzantine or backdoor attacks. To detect
malicious model updates, a typical approach is to measure the distance between
each model update and a \textit{ground-truth model update}. To find such
\textit{ground-truth model updates}, existing defenses either require a benign
root dataset on the server (e.g., FLTrust) or simply use trimmed mean or median
as the threshold for clipping (e.g., FLAME). However, such benign root datasets
are impractical, and the trimmed mean or median may also eliminate
contributions from these underrepresented datasets.
In this paper, we propose a generic solution, namely FedTruth, to defend
against model poisoning attacks in FL, where the \textit{ground-truth model
update} (i.e., the global model update) will be estimated among all the model
updates with dynamic aggregation weights. Specifically, FedTruth does not have
specific assumptions on the benign or malicious data distribution or access to
a benign root dataset. Moreover, FedTruth considers the potential contributions
from all benign clients. Our empirical results show that FedTruth can reduce
the impacts of poisoned model updates against both Byzantine and backdoor
attacks, and is also efficient in large-scale FL systems.
| no_new_dataset | 0.944995 |
2311.14282 | Zheng Chen | Zheng Chen, Yulun Zhang, Jinjin Gu, Xin Yuan, Linghe Kong, Guihai
Chen, Xiaokang Yang | Image Super-Resolution with Text Prompt Diffusion | Code is available at https://github.com/zhengchen1999/PromptSR | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image super-resolution (SR) methods typically model degradation to improve
reconstruction accuracy in complex and unknown degradation scenarios. However,
extracting degradation information from low-resolution images is challenging,
which limits the model performance. To boost image SR performance, one feasible
approach is to introduce additional priors. Inspired by advancements in
multi-modal methods and text prompt image processing, we introduce text prompts
to image SR to provide degradation priors. Specifically, we first design a
text-image generation pipeline to integrate text into the SR dataset through
the text degradation representation and degradation model. By adopting a
discrete design, the text representation is flexible and user-friendly.
Meanwhile, we propose the PromptSR to realize the text prompt SR. The PromptSR
leverages the latest multi-modal large language model (MLLM) to generate
prompts from low-resolution images. It also utilizes the pre-trained language
model (e.g., T5 or CLIP) to enhance text comprehension. We train the PromptSR
on the text-image dataset. Extensive experiments indicate that introducing text
prompts into SR, yields impressive results on both synthetic and real-world
images. Code: https://github.com/zhengchen1999/PromptSR.
| [
{
"version": "v1",
"created": "Fri, 24 Nov 2023 05:11:35 GMT"
},
{
"version": "v2",
"created": "Tue, 12 Mar 2024 12:14:51 GMT"
},
{
"version": "v3",
"created": "Tue, 8 Oct 2024 10:30:00 GMT"
},
{
"version": "v4",
"created": "Thu, 10 Oct 2024 05:47:46 GMT"
},
{
"version": "v5",
"created": "Tue, 11 Mar 2025 02:20:58 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Chen",
"Zheng",
""
],
[
"Zhang",
"Yulun",
""
],
[
"Gu",
"Jinjin",
""
],
[
"Yuan",
"Xin",
""
],
[
"Kong",
"Linghe",
""
],
[
"Chen",
"Guihai",
""
],
[
"Yang",
"Xiaokang",
""
]
]
| TITLE: Image Super-Resolution with Text Prompt Diffusion
ABSTRACT: Image super-resolution (SR) methods typically model degradation to improve
reconstruction accuracy in complex and unknown degradation scenarios. However,
extracting degradation information from low-resolution images is challenging,
which limits the model performance. To boost image SR performance, one feasible
approach is to introduce additional priors. Inspired by advancements in
multi-modal methods and text prompt image processing, we introduce text prompts
to image SR to provide degradation priors. Specifically, we first design a
text-image generation pipeline to integrate text into the SR dataset through
the text degradation representation and degradation model. By adopting a
discrete design, the text representation is flexible and user-friendly.
Meanwhile, we propose the PromptSR to realize the text prompt SR. The PromptSR
leverages the latest multi-modal large language model (MLLM) to generate
prompts from low-resolution images. It also utilizes the pre-trained language
model (e.g., T5 or CLIP) to enhance text comprehension. We train the PromptSR
on the text-image dataset. Extensive experiments indicate that introducing text
prompts into SR, yields impressive results on both synthetic and real-world
images. Code: https://github.com/zhengchen1999/PromptSR.
| no_new_dataset | 0.950365 |
2312.07896 | Joshua Groen | Joshua Groen, Zixian Yang, Divyadharshini Muruganandham, Mauro
Belgiovine, Lei Ying, Kaushik Chowdhury | From Classification to Optimization: Slicing and Resource Management
with TRACTOR | Journal version of TRACTOR: Traffic Analysis and Classification Tool
for Open RAN | null | null | null | eess.SY cs.NI cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 5G and beyond networks promise advancements in bandwidth, latency, and
connectivity. The Open Radio Access Network (O-RAN) framework enhances
flexibility through network slicing and closed-loop RAN control. Central to
this evolution is integrating machine learning (ML) for dynamic network
control. This paper presents a framework to optimize O-RAN operation. First, we
build and share a robust O-RAN dataset from real-world traffic captured across
diverse locations and mobility scenarios, replicated within a full-stack
srsRAN-based O-RAN system using the Colosseum RF emulator. This dataset
supports ML training and deployment. We then introduce a traffic classification
approach leveraging various ML models, demonstrating rapid training, testing,
and refinement to improve accuracy. With up to 99% offline accuracy and 92%
online accuracy for specific slices, our framework adapts efficiently to
different models and network conditions. Finally, we present a physical
resource block (PRB) assignment optimization strategy using reinforcement
learning to refine resource allocation. Our learned policy achieves a mean
performance score (0.631), surpassing a manually configured expert policy
(0.609) and a random baseline (0.588), demonstrating improved PRB utilization.
More importantly, our approach exhibits lower variability, with the Coefficient
of Variation (CV) reduced by up to an order of magnitude in three out of four
cases, ensuring more consistent performance. Our contributions, including
open-source tools and datasets, accelerate O-RAN and ML-driven network control
research.
| [
{
"version": "v1",
"created": "Wed, 13 Dec 2023 04:53:12 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 15:31:55 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Groen",
"Joshua",
""
],
[
"Yang",
"Zixian",
""
],
[
"Muruganandham",
"Divyadharshini",
""
],
[
"Belgiovine",
"Mauro",
""
],
[
"Ying",
"Lei",
""
],
[
"Chowdhury",
"Kaushik",
""
]
]
| TITLE: From Classification to Optimization: Slicing and Resource Management
with TRACTOR
ABSTRACT: 5G and beyond networks promise advancements in bandwidth, latency, and
connectivity. The Open Radio Access Network (O-RAN) framework enhances
flexibility through network slicing and closed-loop RAN control. Central to
this evolution is integrating machine learning (ML) for dynamic network
control. This paper presents a framework to optimize O-RAN operation. First, we
build and share a robust O-RAN dataset from real-world traffic captured across
diverse locations and mobility scenarios, replicated within a full-stack
srsRAN-based O-RAN system using the Colosseum RF emulator. This dataset
supports ML training and deployment. We then introduce a traffic classification
approach leveraging various ML models, demonstrating rapid training, testing,
and refinement to improve accuracy. With up to 99% offline accuracy and 92%
online accuracy for specific slices, our framework adapts efficiently to
different models and network conditions. Finally, we present a physical
resource block (PRB) assignment optimization strategy using reinforcement
learning to refine resource allocation. Our learned policy achieves a mean
performance score (0.631), surpassing a manually configured expert policy
(0.609) and a random baseline (0.588), demonstrating improved PRB utilization.
More importantly, our approach exhibits lower variability, with the Coefficient
of Variation (CV) reduced by up to an order of magnitude in three out of four
cases, ensuring more consistent performance. Our contributions, including
open-source tools and datasets, accelerate O-RAN and ML-driven network control
research.
| no_new_dataset | 0.932269 |
2312.08177 | Peilin Cai | Peilin Cai | Advanced Image Segmentation Techniques for Neural Activity Detection via
C-fos Immediate Early Gene Expression | The paper has major omissions, a factual error in the description of
UNet in the illustration on the second page, and extra printing on the last
page, which is terrible. The original document is also lost and cannot be
modified | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates the application of advanced image segmentation
techniques to analyze C-fos immediate early gene expression, a crucial marker
for neural activity. Due to the complexity and high variability of neural
circuits, accurate segmentation of C-fos images is paramount for the
development of new insights into neural function. Amidst this backdrop, this
research aims to improve accuracy and minimize manual intervention in C-fos
image segmentation by leveraging the capabilities of CNNs and the Unet model.
We describe the development of a novel workflow for the segmentation process
involving Convolutional Neural Networks (CNNs) and the Unet model,
demonstrating their efficiency in various image segmentation tasks. Our
workflow incorporates pre-processing steps such as cropping, image feature
extraction, and clustering for the training dataset selection. We used an
AutoEncoder model to extract features and implement constrained clustering to
identify similarities and differences in image types. Additionally, we utilized
manual and automatic labeling approaches to enhance the performance of our
model. We demonstrated the effectiveness of our method in distinguishing areas
with significant C-fos expression from normal tissue areas. Lastly, we
implemented a modified Unet network for the detection of C-fos expressions.
This research contributes to the development of more efficient and automated
image segmentation methods, advancing the understanding of neural function in
neuroscience research.
| [
{
"version": "v1",
"created": "Wed, 13 Dec 2023 14:36:16 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 23:42:49 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Cai",
"Peilin",
""
]
]
| TITLE: Advanced Image Segmentation Techniques for Neural Activity Detection via
C-fos Immediate Early Gene Expression
ABSTRACT: This paper investigates the application of advanced image segmentation
techniques to analyze C-fos immediate early gene expression, a crucial marker
for neural activity. Due to the complexity and high variability of neural
circuits, accurate segmentation of C-fos images is paramount for the
development of new insights into neural function. Amidst this backdrop, this
research aims to improve accuracy and minimize manual intervention in C-fos
image segmentation by leveraging the capabilities of CNNs and the Unet model.
We describe the development of a novel workflow for the segmentation process
involving Convolutional Neural Networks (CNNs) and the Unet model,
demonstrating their efficiency in various image segmentation tasks. Our
workflow incorporates pre-processing steps such as cropping, image feature
extraction, and clustering for the training dataset selection. We used an
AutoEncoder model to extract features and implement constrained clustering to
identify similarities and differences in image types. Additionally, we utilized
manual and automatic labeling approaches to enhance the performance of our
model. We demonstrated the effectiveness of our method in distinguishing areas
with significant C-fos expression from normal tissue areas. Lastly, we
implemented a modified Unet network for the detection of C-fos expressions.
This research contributes to the development of more efficient and automated
image segmentation methods, advancing the understanding of neural function in
neuroscience research.
| no_new_dataset | 0.952926 |
2401.00740 | Zeke Zexi Hu | Zeke Zexi Hu, Xiaoming Chen, Vera Yuk Ying Chung, Yiran Shen | Beyond Subspace Isolation: Many-to-Many Transformer for Light Field
Image Super-resolution | Accepted by IEEE Transactions on Multimedia | null | 10.1109/TMM.2024.3521795 | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The effective extraction of spatial-angular features plays a crucial role in
light field image super-resolution (LFSR) tasks, and the introduction of
convolution and Transformers leads to significant improvement in this area.
Nevertheless, due to the large 4D data volume of light field images, many
existing methods opted to decompose the data into a number of lower-dimensional
subspaces and perform Transformers in each sub-space individually. As a side
effect, these methods inadvertently restrict the self-attention mechanisms to a
One-to-One scheme accessing only a limited subset of LF data, explicitly
preventing comprehensive optimization on all spatial and angular cues. In this
paper, we identify this limitation as subspace isolation and introduce a novel
Many-to-Many Transformer (M2MT) to address it. M2MT aggregates angular
information in the spatial subspace before performing the self-attention
mechanism. It enables complete access to all information across all
sub-aperture images (SAIs) in a light field image. Consequently, M2MT is
enabled to comprehensively capture long-range correlation dependencies. With
M2MT as the pivotal component, we develop a simple yet effective M2MT network
for LFSR. Our experimental results demonstrate that M2MT achieves
state-of-the-art performance across various public datasets. We further conduct
in-depth analysis using local attribution maps (LAM) to obtain visual
interpretability, and the results validate that M2MT is empowered with a truly
non-local context in both spatial and angular subspaces to mitigate subspace
isolation and acquire effective spatial-angular representation.
| [
{
"version": "v1",
"created": "Mon, 1 Jan 2024 12:48:23 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 12:54:24 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Hu",
"Zeke Zexi",
""
],
[
"Chen",
"Xiaoming",
""
],
[
"Chung",
"Vera Yuk Ying",
""
],
[
"Shen",
"Yiran",
""
]
]
| TITLE: Beyond Subspace Isolation: Many-to-Many Transformer for Light Field
Image Super-resolution
ABSTRACT: The effective extraction of spatial-angular features plays a crucial role in
light field image super-resolution (LFSR) tasks, and the introduction of
convolution and Transformers leads to significant improvement in this area.
Nevertheless, due to the large 4D data volume of light field images, many
existing methods opted to decompose the data into a number of lower-dimensional
subspaces and perform Transformers in each sub-space individually. As a side
effect, these methods inadvertently restrict the self-attention mechanisms to a
One-to-One scheme accessing only a limited subset of LF data, explicitly
preventing comprehensive optimization on all spatial and angular cues. In this
paper, we identify this limitation as subspace isolation and introduce a novel
Many-to-Many Transformer (M2MT) to address it. M2MT aggregates angular
information in the spatial subspace before performing the self-attention
mechanism. It enables complete access to all information across all
sub-aperture images (SAIs) in a light field image. Consequently, M2MT is
enabled to comprehensively capture long-range correlation dependencies. With
M2MT as the pivotal component, we develop a simple yet effective M2MT network
for LFSR. Our experimental results demonstrate that M2MT achieves
state-of-the-art performance across various public datasets. We further conduct
in-depth analysis using local attribution maps (LAM) to obtain visual
interpretability, and the results validate that M2MT is empowered with a truly
non-local context in both spatial and angular subspaces to mitigate subspace
isolation and acquire effective spatial-angular representation.
| no_new_dataset | 0.950457 |
2401.01759 | Lin Bai | Lin Bai, Caiyan Jia, Ziying Song, and Chaoqun Cui | VGA: Vision and Graph Fused Attention Network for Rumor Detection | null | null | 10.1145/3722225 | null | cs.SI cs.CL cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the development of social media, rumors have been spread broadly on
social media platforms, causing great harm to society. Beside textual
information, many rumors also use manipulated images or conceal textual
information within images to deceive people and avoid being detected, making
multimodal rumor detection be a critical problem. The majority of multimodal
rumor detection methods mainly concentrate on extracting features of source
claims and their corresponding images, while ignoring the comments of rumors
and their propagation structures. These comments and structures imply the
wisdom of crowds and are proved to be crucial to debunk rumors. Moreover, these
methods usually only extract visual features in a basic manner, seldom consider
tampering or textual information in images. Therefore, in this study, we
propose a novel Vision and Graph Fused Attention Network (VGA) for rumor
detection to utilize propagation structures among posts so as to obtain the
crowd opinions and further explore visual tampering features, as well as the
textual information hidden in images. We conduct extensive experiments on three
datasets, demonstrating that VGA can effectively detect multimodal rumors and
outperform state-of-the-art methods significantly.
| [
{
"version": "v1",
"created": "Wed, 3 Jan 2024 14:24:02 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Bai",
"Lin",
""
],
[
"Jia",
"Caiyan",
""
],
[
"Song",
"Ziying",
""
],
[
"Cui",
"Chaoqun",
""
]
]
| TITLE: VGA: Vision and Graph Fused Attention Network for Rumor Detection
ABSTRACT: With the development of social media, rumors have been spread broadly on
social media platforms, causing great harm to society. Beside textual
information, many rumors also use manipulated images or conceal textual
information within images to deceive people and avoid being detected, making
multimodal rumor detection be a critical problem. The majority of multimodal
rumor detection methods mainly concentrate on extracting features of source
claims and their corresponding images, while ignoring the comments of rumors
and their propagation structures. These comments and structures imply the
wisdom of crowds and are proved to be crucial to debunk rumors. Moreover, these
methods usually only extract visual features in a basic manner, seldom consider
tampering or textual information in images. Therefore, in this study, we
propose a novel Vision and Graph Fused Attention Network (VGA) for rumor
detection to utilize propagation structures among posts so as to obtain the
crowd opinions and further explore visual tampering features, as well as the
textual information hidden in images. We conduct extensive experiments on three
datasets, demonstrating that VGA can effectively detect multimodal rumors and
outperform state-of-the-art methods significantly.
| no_new_dataset | 0.94743 |
2402.01974 | LIanhao Yin | Lianhao Yin, Yutong Ban, Jennifer Eckhoff, Ozanan Meireles, Daniela
Rus, Guy Rosman | Hypergraph-Transformer (HGT) for Interactive Event Prediction in
Laparoscopic and Robotic Surgery | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding and anticipating intraoperative events and actions is critical
for intraoperative assistance and decision-making during minimally invasive
surgery. Automated prediction of events, actions, and the following
consequences is addressed through various computational approaches with the
objective of augmenting surgeons' perception and decision-making capabilities.
We propose a predictive neural network that is capable of understanding and
predicting critical interactive aspects of surgical workflow from
intra-abdominal video, while flexibly leveraging surgical knowledge graphs. The
approach incorporates a hypergraph-transformer (HGT) structure that encodes
expert knowledge into the network design and predicts the hidden embedding of
the graph. We verify our approach on established surgical datasets and
applications, including the detection and prediction of action triplets, and
the achievement of the Critical View of Safety (CVS). Moreover, we address
specific, safety-related tasks, such as predicting the clipping of cystic duct
or artery without prior achievement of the CVS. Our results demonstrate the
superiority of our approach compared to unstructured alternatives.
| [
{
"version": "v1",
"created": "Sat, 3 Feb 2024 00:58:05 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 21:58:42 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Yin",
"Lianhao",
""
],
[
"Ban",
"Yutong",
""
],
[
"Eckhoff",
"Jennifer",
""
],
[
"Meireles",
"Ozanan",
""
],
[
"Rus",
"Daniela",
""
],
[
"Rosman",
"Guy",
""
]
]
| TITLE: Hypergraph-Transformer (HGT) for Interactive Event Prediction in
Laparoscopic and Robotic Surgery
ABSTRACT: Understanding and anticipating intraoperative events and actions is critical
for intraoperative assistance and decision-making during minimally invasive
surgery. Automated prediction of events, actions, and the following
consequences is addressed through various computational approaches with the
objective of augmenting surgeons' perception and decision-making capabilities.
We propose a predictive neural network that is capable of understanding and
predicting critical interactive aspects of surgical workflow from
intra-abdominal video, while flexibly leveraging surgical knowledge graphs. The
approach incorporates a hypergraph-transformer (HGT) structure that encodes
expert knowledge into the network design and predicts the hidden embedding of
the graph. We verify our approach on established surgical datasets and
applications, including the detection and prediction of action triplets, and
the achievement of the Critical View of Safety (CVS). Moreover, we address
specific, safety-related tasks, such as predicting the clipping of cystic duct
or artery without prior achievement of the CVS. Our results demonstrate the
superiority of our approach compared to unstructured alternatives.
| no_new_dataset | 0.944074 |
2402.04412 | Andrew Stirn | Andrew A. Stirn and David A. Knowles | The VampPrior Mixture Model | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Widely used deep latent variable models (DLVMs), in particular Variational
Autoencoders (VAEs), employ overly simplistic priors on the latent space. To
achieve strong clustering performance, existing methods that replace the
standard normal prior with a Gaussian mixture model (GMM) require defining the
number of clusters to be close to the number of expected ground truth classes
a-priori and are susceptible to poor initializations. We leverage VampPrior
concepts (Tomczak and Welling, 2018) to fit a Bayesian GMM prior, resulting in
the VampPrior Mixture Model (VMM), a novel prior for DLVMs. In a VAE, the VMM
attains highly competitive clustering performance on benchmark datasets.
Integrating the VMM into scVI (Lopez et al., 2018), a popular scRNA-seq
integration method, significantly improves its performance and automatically
arranges cells into clusters with similar biological characteristics.
| [
{
"version": "v1",
"created": "Tue, 6 Feb 2024 21:18:34 GMT"
},
{
"version": "v2",
"created": "Tue, 5 Mar 2024 03:52:18 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 03:56:24 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Stirn",
"Andrew A.",
""
],
[
"Knowles",
"David A.",
""
]
]
| TITLE: The VampPrior Mixture Model
ABSTRACT: Widely used deep latent variable models (DLVMs), in particular Variational
Autoencoders (VAEs), employ overly simplistic priors on the latent space. To
achieve strong clustering performance, existing methods that replace the
standard normal prior with a Gaussian mixture model (GMM) require defining the
number of clusters to be close to the number of expected ground truth classes
a-priori and are susceptible to poor initializations. We leverage VampPrior
concepts (Tomczak and Welling, 2018) to fit a Bayesian GMM prior, resulting in
the VampPrior Mixture Model (VMM), a novel prior for DLVMs. In a VAE, the VMM
attains highly competitive clustering performance on benchmark datasets.
Integrating the VMM into scVI (Lopez et al., 2018), a popular scRNA-seq
integration method, significantly improves its performance and automatically
arranges cells into clusters with similar biological characteristics.
| no_new_dataset | 0.948775 |
2403.09905 | Vishnu Sashank Dorbala | Vishnu Sashank Dorbala, Bhrij Patel, Amrit Singh Bedi, Dinesh Manocha | Right Place, Right Time! Dynamizing Topological Graphs for Embodied
Navigation | 18 | null | null | null | cs.RO cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | Embodied Navigation tasks often involve constructing topological graphs of a
scene during exploration to facilitate high-level planning and decision-making
for execution in continuous environments. Prior literature makes the assumption
of static graphs with stationary targets, which does not hold in many
real-world environments with moving objects. To address this, we present a
novel formulation generalizing navigation to dynamic environments by
introducing structured object transitions to dynamize static topological graphs
called Object Transition Graphs (OTGs). OTGs simulate portable targets
following structured routes inspired by human habits. We apply this technique
to Matterport3D (MP3D), a popular simulator for evaluating embodied tasks. On
these dynamized OTGs, we establish a navigation benchmark by evaluating
Oracle-based, Reinforcement Learning, and Large Language Model (LLM)-based
approaches on a multi-object finding task. Further, we quantify agent
adaptability, and make key inferences such as agents employing learned
decision-making strategies generalize better than those relying on privileged
oracle knowledge. To the best of our knowledge, ours is the first work to
introduce structured temporal dynamism on topological graphs for studying
generalist embodied navigation policies. The code and dataset for our OTGs will
be made publicly available to foster research on embodied navigation in dynamic
scenes.
| [
{
"version": "v1",
"created": "Thu, 14 Mar 2024 22:33:22 GMT"
},
{
"version": "v2",
"created": "Sun, 1 Dec 2024 21:42:37 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Mar 2025 22:26:37 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Dorbala",
"Vishnu Sashank",
""
],
[
"Patel",
"Bhrij",
""
],
[
"Bedi",
"Amrit Singh",
""
],
[
"Manocha",
"Dinesh",
""
]
]
| TITLE: Right Place, Right Time! Dynamizing Topological Graphs for Embodied
Navigation
ABSTRACT: Embodied Navigation tasks often involve constructing topological graphs of a
scene during exploration to facilitate high-level planning and decision-making
for execution in continuous environments. Prior literature makes the assumption
of static graphs with stationary targets, which does not hold in many
real-world environments with moving objects. To address this, we present a
novel formulation generalizing navigation to dynamic environments by
introducing structured object transitions to dynamize static topological graphs
called Object Transition Graphs (OTGs). OTGs simulate portable targets
following structured routes inspired by human habits. We apply this technique
to Matterport3D (MP3D), a popular simulator for evaluating embodied tasks. On
these dynamized OTGs, we establish a navigation benchmark by evaluating
Oracle-based, Reinforcement Learning, and Large Language Model (LLM)-based
approaches on a multi-object finding task. Further, we quantify agent
adaptability, and make key inferences such as agents employing learned
decision-making strategies generalize better than those relying on privileged
oracle knowledge. To the best of our knowledge, ours is the first work to
introduce structured temporal dynamism on topological graphs for studying
generalist embodied navigation policies. The code and dataset for our OTGs will
be made publicly available to foster research on embodied navigation in dynamic
scenes.
| no_new_dataset | 0.871803 |
2404.02611 | Ivan Sevillano-Garc\'ia | Iv\'an Sevillano-Garc\'ia, Juli\'an Luengo and Francisco Herrera | X-SHIELD: Regularization for eXplainable Artificial Intelligence | 18 pages, 9 figures | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | As artificial intelligence systems become integral across domains, the demand
for explainability grows, the called eXplainable artificial intelligence (XAI).
Existing efforts primarily focus on generating and evaluating explanations for
black-box models while a critical gap in directly enhancing models remains
through these evaluations. It is important to consider the potential of this
explanation process to improve model quality with a feedback on training as
well. XAI may be used to improve model performance while boosting its
explainability. Under this view, this paper introduces Transformation -
Selective Hidden Input Evaluation for Learning Dynamics (T-SHIELD), a
regularization family designed to improve model quality by hiding features of
input, forcing the model to generalize without those features. Within this
family, we propose the XAI - SHIELD(X-SHIELD), a regularization for explainable
artificial intelligence, which uses explanations to select specific features to
hide. In contrast to conventional approaches, X-SHIELD regularization
seamlessly integrates into the objective function enhancing model
explainability while also improving performance. Experimental validation on
benchmark datasets underscores X-SHIELD's effectiveness in improving
performance and overall explainability. The improvement is validated through
experiments comparing models with and without the X-SHIELD regularization, with
further analysis exploring the rationale behind its design choices. This
establishes X-SHIELD regularization as a promising pathway for developing
reliable artificial intelligence regularization.
| [
{
"version": "v1",
"created": "Wed, 3 Apr 2024 09:56:38 GMT"
},
{
"version": "v2",
"created": "Thu, 14 Nov 2024 22:53:12 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 12:24:01 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Sevillano-García",
"Iván",
""
],
[
"Luengo",
"Julián",
""
],
[
"Herrera",
"Francisco",
""
]
]
| TITLE: X-SHIELD: Regularization for eXplainable Artificial Intelligence
ABSTRACT: As artificial intelligence systems become integral across domains, the demand
for explainability grows, the called eXplainable artificial intelligence (XAI).
Existing efforts primarily focus on generating and evaluating explanations for
black-box models while a critical gap in directly enhancing models remains
through these evaluations. It is important to consider the potential of this
explanation process to improve model quality with a feedback on training as
well. XAI may be used to improve model performance while boosting its
explainability. Under this view, this paper introduces Transformation -
Selective Hidden Input Evaluation for Learning Dynamics (T-SHIELD), a
regularization family designed to improve model quality by hiding features of
input, forcing the model to generalize without those features. Within this
family, we propose the XAI - SHIELD(X-SHIELD), a regularization for explainable
artificial intelligence, which uses explanations to select specific features to
hide. In contrast to conventional approaches, X-SHIELD regularization
seamlessly integrates into the objective function enhancing model
explainability while also improving performance. Experimental validation on
benchmark datasets underscores X-SHIELD's effectiveness in improving
performance and overall explainability. The improvement is validated through
experiments comparing models with and without the X-SHIELD regularization, with
further analysis exploring the rationale behind its design choices. This
establishes X-SHIELD regularization as a promising pathway for developing
reliable artificial intelligence regularization.
| no_new_dataset | 0.941331 |
2404.10419 | Matthieu Futeral | Matthieu Futeral, Andrea Agostinelli, Marco Tagliasacchi, Neil
Zeghidour, Eugene Kharitonov | MAD Speech: Measures of Acoustic Diversity of Speech | NAACL 2025 | null | null | null | eess.AS cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative spoken language models produce speech in a wide range of voices,
prosody, and recording conditions, seemingly approaching the diversity of
natural speech. However, the extent to which generated speech is acoustically
diverse remains unclear due to a lack of appropriate metrics. We address this
gap by developing lightweight metrics of acoustic diversity, which we
collectively refer to as MAD Speech. We focus on measuring five facets of
acoustic diversity: voice, gender, emotion, accent, and background noise. We
construct the metrics as a composition of specialized, per-facet embedding
models and an aggregation function that measures diversity within the embedding
space. Next, we build a series of datasets with a priori known diversity
preferences for each facet. Using these datasets, we demonstrate that our
proposed metrics achieve a stronger agreement with the ground-truth diversity
than baselines. Finally, we showcase the applicability of our proposed metrics
across several real-life evaluation scenarios. MAD Speech is made publicly
accessible.
| [
{
"version": "v1",
"created": "Tue, 16 Apr 2024 09:35:27 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 12:02:06 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Futeral",
"Matthieu",
""
],
[
"Agostinelli",
"Andrea",
""
],
[
"Tagliasacchi",
"Marco",
""
],
[
"Zeghidour",
"Neil",
""
],
[
"Kharitonov",
"Eugene",
""
]
]
| TITLE: MAD Speech: Measures of Acoustic Diversity of Speech
ABSTRACT: Generative spoken language models produce speech in a wide range of voices,
prosody, and recording conditions, seemingly approaching the diversity of
natural speech. However, the extent to which generated speech is acoustically
diverse remains unclear due to a lack of appropriate metrics. We address this
gap by developing lightweight metrics of acoustic diversity, which we
collectively refer to as MAD Speech. We focus on measuring five facets of
acoustic diversity: voice, gender, emotion, accent, and background noise. We
construct the metrics as a composition of specialized, per-facet embedding
models and an aggregation function that measures diversity within the embedding
space. Next, we build a series of datasets with a priori known diversity
preferences for each facet. Using these datasets, we demonstrate that our
proposed metrics achieve a stronger agreement with the ground-truth diversity
than baselines. Finally, we showcase the applicability of our proposed metrics
across several real-life evaluation scenarios. MAD Speech is made publicly
accessible.
| new_dataset | 0.958577 |
2404.11868 | Azad Singh | Vandan Gorade, Azad Singh, and Deepak Mishra | OTCXR: Rethinking Self-supervised Alignment using Optimal Transport for
Chest X-ray Analysis | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-supervised learning (SSL) has emerged as a promising technique for
analyzing medical modalities such as X-rays due to its ability to learn without
annotations. However, conventional SSL methods face challenges in achieving
semantic alignment and capturing subtle details, which limits their ability to
accurately represent the underlying anatomical structures and pathological
features. To address these limitations, we propose OTCXR, a novel SSL framework
that leverages optimal transport (OT) to learn dense semantic invariance. By
integrating OT with our innovative Cross-Viewpoint Semantics Infusion Module
(CV-SIM), OTCXR enhances the model's ability to capture not only local spatial
features but also global contextual dependencies across different viewpoints.
This approach enriches the effectiveness of SSL in the context of chest
radiographs. Furthermore, OTCXR incorporates variance and covariance
regularizations within the OT framework to prioritize clinically relevant
information while suppressing less informative features. This ensures that the
learned representations are comprehensive and discriminative, particularly
beneficial for tasks such as thoracic disease diagnosis. We validate OTCXR's
efficacy through comprehensive experiments on three publicly available chest
X-ray datasets. Our empirical results demonstrate the superiority of OTCXR over
state-of-the-art methods across all evaluated tasks, confirming its capability
to learn semantically rich representations.
| [
{
"version": "v1",
"created": "Thu, 18 Apr 2024 02:59:48 GMT"
},
{
"version": "v2",
"created": "Wed, 24 Apr 2024 06:05:33 GMT"
},
{
"version": "v3",
"created": "Sun, 12 May 2024 03:15:07 GMT"
},
{
"version": "v4",
"created": "Tue, 11 Mar 2025 10:09:11 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Gorade",
"Vandan",
""
],
[
"Singh",
"Azad",
""
],
[
"Mishra",
"Deepak",
""
]
]
| TITLE: OTCXR: Rethinking Self-supervised Alignment using Optimal Transport for
Chest X-ray Analysis
ABSTRACT: Self-supervised learning (SSL) has emerged as a promising technique for
analyzing medical modalities such as X-rays due to its ability to learn without
annotations. However, conventional SSL methods face challenges in achieving
semantic alignment and capturing subtle details, which limits their ability to
accurately represent the underlying anatomical structures and pathological
features. To address these limitations, we propose OTCXR, a novel SSL framework
that leverages optimal transport (OT) to learn dense semantic invariance. By
integrating OT with our innovative Cross-Viewpoint Semantics Infusion Module
(CV-SIM), OTCXR enhances the model's ability to capture not only local spatial
features but also global contextual dependencies across different viewpoints.
This approach enriches the effectiveness of SSL in the context of chest
radiographs. Furthermore, OTCXR incorporates variance and covariance
regularizations within the OT framework to prioritize clinically relevant
information while suppressing less informative features. This ensures that the
learned representations are comprehensive and discriminative, particularly
beneficial for tasks such as thoracic disease diagnosis. We validate OTCXR's
efficacy through comprehensive experiments on three publicly available chest
X-ray datasets. Our empirical results demonstrate the superiority of OTCXR over
state-of-the-art methods across all evaluated tasks, confirming its capability
to learn semantically rich representations.
| no_new_dataset | 0.946349 |
2404.17884 | Rodrigo Abad\'ia Heredia | Rodrigo Abad\'ia-Heredia, Adri\'an Corrochano, Manuel Lopez-Martin,
Soledad Le Clainche | Generalization capabilities and robustness of hybrid models grounded in
physics compared to purely deep learning models | 24 pages, two column, 26 figures and 11 tables | Physics of Fluids 1 March 2025; 37 (3): 035149 | 10.1063/5.0253876 | null | physics.flu-dyn cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This study investigates the generalization capabilities and robustness of
purely deep learning (DL) models and hybrid models based on physical principles
in fluid dynamics applications, specifically focusing on iteratively
forecasting the temporal evolution of flow dynamics. Three autoregressive
models were compared: a hybrid model (POD-DL) that combines proper orthogonal
decomposition (POD) with a long-short term memory (LSTM) layer, a convolutional
autoencoder combined with a convolutional LSTM (ConvLSTM) layer and a
variational autoencoder (VAE) combined with a ConvLSTM layer. These models were
tested on two high-dimensional, nonlinear datasets representing the velocity
field of flow past a circular cylinder in both laminar and turbulent regimes.
The study used latent dimension methods, enabling a bijective reduction of
high-dimensional dynamics into a lower-order space to facilitate future
predictions. While the VAE and ConvLSTM models accurately predicted laminar
flow, the hybrid POD-DL model outperformed the others across both laminar and
turbulent flow regimes. This success is attributed to the model's ability to
incorporate modal decomposition, reducing the dimensionality of the data, by a
non-parametric method, and simplifying the forecasting component. By leveraging
POD, the model not only gained insight into the underlying physics, improving
prediction accuracy with less training data, but also reduce the number of
trainable parameters as POD is non-parametric. The findings emphasize the
potential of hybrid models, particularly those integrating modal decomposition
and deep learning, in predicting complex flow dynamics.
| [
{
"version": "v1",
"created": "Sat, 27 Apr 2024 12:43:02 GMT"
},
{
"version": "v2",
"created": "Mon, 28 Oct 2024 15:31:11 GMT"
},
{
"version": "v3",
"created": "Fri, 24 Jan 2025 19:50:59 GMT"
},
{
"version": "v4",
"created": "Mon, 17 Feb 2025 15:37:58 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Abadía-Heredia",
"Rodrigo",
""
],
[
"Corrochano",
"Adrián",
""
],
[
"Lopez-Martin",
"Manuel",
""
],
[
"Clainche",
"Soledad Le",
""
]
]
| TITLE: Generalization capabilities and robustness of hybrid models grounded in
physics compared to purely deep learning models
ABSTRACT: This study investigates the generalization capabilities and robustness of
purely deep learning (DL) models and hybrid models based on physical principles
in fluid dynamics applications, specifically focusing on iteratively
forecasting the temporal evolution of flow dynamics. Three autoregressive
models were compared: a hybrid model (POD-DL) that combines proper orthogonal
decomposition (POD) with a long-short term memory (LSTM) layer, a convolutional
autoencoder combined with a convolutional LSTM (ConvLSTM) layer and a
variational autoencoder (VAE) combined with a ConvLSTM layer. These models were
tested on two high-dimensional, nonlinear datasets representing the velocity
field of flow past a circular cylinder in both laminar and turbulent regimes.
The study used latent dimension methods, enabling a bijective reduction of
high-dimensional dynamics into a lower-order space to facilitate future
predictions. While the VAE and ConvLSTM models accurately predicted laminar
flow, the hybrid POD-DL model outperformed the others across both laminar and
turbulent flow regimes. This success is attributed to the model's ability to
incorporate modal decomposition, reducing the dimensionality of the data, by a
non-parametric method, and simplifying the forecasting component. By leveraging
POD, the model not only gained insight into the underlying physics, improving
prediction accuracy with less training data, but also reduce the number of
trainable parameters as POD is non-parametric. The findings emphasize the
potential of hybrid models, particularly those integrating modal decomposition
and deep learning, in predicting complex flow dynamics.
| no_new_dataset | 0.949482 |
2406.03486 | Soonwoo Kwon | Soonwoo Kwon, Sojung Kim, Minju Park, Seunghyun Lee, Kyuseok Kim | BIPED: Pedagogically Informed Tutoring System for ESL Education | ACL 2024 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) have a great potential to serve as readily
available and cost-efficient Conversational Intelligent Tutoring Systems (CITS)
for teaching L2 learners of English. Existing CITS, however, are designed to
teach only simple concepts or lack the pedagogical depth necessary to address
diverse learning strategies. To develop a more pedagogically informed CITS
capable of teaching complex concepts, we construct a BIlingual
PEDagogically-informed Tutoring Dataset (BIPED) of one-on-one, human-to-human
English tutoring interactions. Through post-hoc analysis of the tutoring
interactions, we come up with a lexicon of dialogue acts (34 tutor acts and 9
student acts), which we use to further annotate the collected dataset. Based on
a two-step framework of first predicting the appropriate tutor act then
generating the corresponding response, we implemented two CITS models using
GPT-4 and SOLAR-KO, respectively. We experimentally demonstrate that the
implemented models not only replicate the style of human teachers but also
employ diverse and contextually appropriate pedagogical strategies.
| [
{
"version": "v1",
"created": "Wed, 5 Jun 2024 17:49:24 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Kwon",
"Soonwoo",
""
],
[
"Kim",
"Sojung",
""
],
[
"Park",
"Minju",
""
],
[
"Lee",
"Seunghyun",
""
],
[
"Kim",
"Kyuseok",
""
]
]
| TITLE: BIPED: Pedagogically Informed Tutoring System for ESL Education
ABSTRACT: Large Language Models (LLMs) have a great potential to serve as readily
available and cost-efficient Conversational Intelligent Tutoring Systems (CITS)
for teaching L2 learners of English. Existing CITS, however, are designed to
teach only simple concepts or lack the pedagogical depth necessary to address
diverse learning strategies. To develop a more pedagogically informed CITS
capable of teaching complex concepts, we construct a BIlingual
PEDagogically-informed Tutoring Dataset (BIPED) of one-on-one, human-to-human
English tutoring interactions. Through post-hoc analysis of the tutoring
interactions, we come up with a lexicon of dialogue acts (34 tutor acts and 9
student acts), which we use to further annotate the collected dataset. Based on
a two-step framework of first predicting the appropriate tutor act then
generating the corresponding response, we implemented two CITS models using
GPT-4 and SOLAR-KO, respectively. We experimentally demonstrate that the
implemented models not only replicate the style of human teachers but also
employ diverse and contextually appropriate pedagogical strategies.
| new_dataset | 0.95995 |
2406.06843 | Jikai Wang | Jikai Wang, Qifan Zhang, Yu-Wei Chao, Bowen Wen, Xiaohu Guo, Yu Xiang | HO-Cap: A Capture System and Dataset for 3D Reconstruction and Pose
Tracking of Hand-Object Interaction | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a data capture system and a new dataset, HO-Cap, for 3D
reconstruction and pose tracking of hands and objects in videos. The system
leverages multiple RGBD cameras and a HoloLens headset for data collection,
avoiding the use of expensive 3D scanners or mocap systems. We propose a
semi-automatic method for annotating the shape and pose of hands and objects in
the collected videos, significantly reducing the annotation time compared to
manual labeling. With this system, we captured a video dataset of humans
interacting with objects to perform various tasks, including simple
pick-and-place actions, handovers between hands, and using objects according to
their affordance, which can serve as human demonstrations for research in
embodied AI and robot manipulation. Our data capture setup and annotation
framework will be available for the community to use in reconstructing 3D
shapes of objects and human hands and tracking their poses in videos.
| [
{
"version": "v1",
"created": "Mon, 10 Jun 2024 23:25:19 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Jun 2024 20:51:53 GMT"
},
{
"version": "v3",
"created": "Wed, 4 Dec 2024 21:05:00 GMT"
},
{
"version": "v4",
"created": "Tue, 11 Mar 2025 16:48:26 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Wang",
"Jikai",
""
],
[
"Zhang",
"Qifan",
""
],
[
"Chao",
"Yu-Wei",
""
],
[
"Wen",
"Bowen",
""
],
[
"Guo",
"Xiaohu",
""
],
[
"Xiang",
"Yu",
""
]
]
| TITLE: HO-Cap: A Capture System and Dataset for 3D Reconstruction and Pose
Tracking of Hand-Object Interaction
ABSTRACT: We introduce a data capture system and a new dataset, HO-Cap, for 3D
reconstruction and pose tracking of hands and objects in videos. The system
leverages multiple RGBD cameras and a HoloLens headset for data collection,
avoiding the use of expensive 3D scanners or mocap systems. We propose a
semi-automatic method for annotating the shape and pose of hands and objects in
the collected videos, significantly reducing the annotation time compared to
manual labeling. With this system, we captured a video dataset of humans
interacting with objects to perform various tasks, including simple
pick-and-place actions, handovers between hands, and using objects according to
their affordance, which can serve as human demonstrations for research in
embodied AI and robot manipulation. Our data capture setup and annotation
framework will be available for the community to use in reconstructing 3D
shapes of objects and human hands and tracking their poses in videos.
| new_dataset | 0.953535 |
2406.07451 | Xiaoyan Hu | Xiaoyan Hu, Ho-fung Leung, and Farzan Farnia | A Multi-Armed Bandit Approach to Online Selection and Evaluation of
Generative Models | arXiv version | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Existing frameworks for evaluating and comparing generative models consider
an offline setting, where the evaluator has access to large batches of data
produced by the models. However, in practical scenarios, the goal is often to
identify and select the best model using the fewest possible generated samples
to minimize the costs of querying data from the sub-optimal models. In this
work, we propose an online evaluation and selection framework to find the
generative model that maximizes a standard assessment score among a group of
available models. We view the task as a multi-armed bandit (MAB) and propose
upper confidence bound (UCB) bandit algorithms to identify the model producing
data with the best evaluation score that quantifies the quality and diversity
of generated data. Specifically, we develop the MAB-based selection of
generative models considering the Fr\'echet Distance (FD) and Inception Score
(IS) metrics, resulting in the FD-UCB and IS-UCB algorithms. We prove regret
bounds for these algorithms and present numerical results on standard image
datasets. Our empirical results suggest the efficacy of MAB approaches for the
sample-efficient evaluation and selection of deep generative models. The
project code is available at https://github.com/yannxiaoyanhu/dgm-online-eval.
| [
{
"version": "v1",
"created": "Tue, 11 Jun 2024 16:57:48 GMT"
},
{
"version": "v2",
"created": "Thu, 31 Oct 2024 16:48:40 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 10:55:52 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Hu",
"Xiaoyan",
""
],
[
"Leung",
"Ho-fung",
""
],
[
"Farnia",
"Farzan",
""
]
]
| TITLE: A Multi-Armed Bandit Approach to Online Selection and Evaluation of
Generative Models
ABSTRACT: Existing frameworks for evaluating and comparing generative models consider
an offline setting, where the evaluator has access to large batches of data
produced by the models. However, in practical scenarios, the goal is often to
identify and select the best model using the fewest possible generated samples
to minimize the costs of querying data from the sub-optimal models. In this
work, we propose an online evaluation and selection framework to find the
generative model that maximizes a standard assessment score among a group of
available models. We view the task as a multi-armed bandit (MAB) and propose
upper confidence bound (UCB) bandit algorithms to identify the model producing
data with the best evaluation score that quantifies the quality and diversity
of generated data. Specifically, we develop the MAB-based selection of
generative models considering the Fr\'echet Distance (FD) and Inception Score
(IS) metrics, resulting in the FD-UCB and IS-UCB algorithms. We prove regret
bounds for these algorithms and present numerical results on standard image
datasets. Our empirical results suggest the efficacy of MAB approaches for the
sample-efficient evaluation and selection of deep generative models. The
project code is available at https://github.com/yannxiaoyanhu/dgm-online-eval.
| no_new_dataset | 0.949529 |
2406.10118 | Holy Lovenia | Holy Lovenia, Rahmad Mahendra, Salsabil Maulana Akbar, Lester James V.
Miranda, Jennifer Santoso, Elyanah Aco, Akhdan Fadhilah, Jonibek Mansurov,
Joseph Marvin Imperial, Onno P. Kampman, Joel Ruben Antony Moniz, Muhammad
Ravi Shulthan Habibi, Frederikus Hudi, Railey Montalan, Ryan Ignatius,
Joanito Agili Lopo, William Nixon, B\"orje F. Karlsson, James Jaya, Ryandito
Diandaru, Yuze Gao, Patrick Amadeus, Bin Wang, Jan Christian Blaise Cruz,
Chenxi Whitehouse, Ivan Halim Parmonangan, Maria Khelli, Wenyu Zhang, Lucky
Susanto, Reynard Adha Ryanda, Sonny Lazuardi Hermawan, Dan John Velasco,
Muhammad Dehan Al Kautsar, Willy Fitra Hendria, Yasmin Moslem, Noah Flynn,
Muhammad Farid Adilazuarda, Haochen Li, Johanes Lee, R. Damanhuri, Shuo Sun,
Muhammad Reza Qorib, Amirbek Djanibekov, Wei Qi Leong, Quyet V. Do, Niklas
Muennighoff, Tanrada Pansuwan, Ilham Firdausi Putra, Yan Xu, Ngee Chia Tai,
Ayu Purwarianti, Sebastian Ruder, William Tjhi, Peerat Limkonchotiwat, Alham
Fikri Aji, Sedrick Keh, Genta Indra Winata, Ruochen Zhang, Fajri Koto,
Zheng-Xin Yong, Samuel Cahyawijaya | SEACrowd: A Multilingual Multimodal Data Hub and Benchmark Suite for
Southeast Asian Languages | https://seacrowd.github.io/ Published in EMNLP 2024 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Southeast Asia (SEA) is a region rich in linguistic diversity and cultural
variety, with over 1,300 indigenous languages and a population of 671 million
people. However, prevailing AI models suffer from a significant lack of
representation of texts, images, and audio datasets from SEA, compromising the
quality of AI models for SEA languages. Evaluating models for SEA languages is
challenging due to the scarcity of high-quality datasets, compounded by the
dominance of English training data, raising concerns about potential cultural
misrepresentation. To address these challenges, we introduce SEACrowd, a
collaborative initiative that consolidates a comprehensive resource hub that
fills the resource gap by providing standardized corpora in nearly 1,000 SEA
languages across three modalities. Through our SEACrowd benchmarks, we assess
the quality of AI models on 36 indigenous languages across 13 tasks, offering
valuable insights into the current AI landscape in SEA. Furthermore, we propose
strategies to facilitate greater AI advancements, maximizing potential utility
and resource equity for the future of AI in SEA.
| [
{
"version": "v1",
"created": "Fri, 14 Jun 2024 15:23:39 GMT"
},
{
"version": "v2",
"created": "Fri, 5 Jul 2024 05:28:20 GMT"
},
{
"version": "v3",
"created": "Mon, 8 Jul 2024 07:49:40 GMT"
},
{
"version": "v4",
"created": "Tue, 8 Oct 2024 14:35:36 GMT"
},
{
"version": "v5",
"created": "Tue, 11 Mar 2025 02:04:36 GMT"
}
]
| 2025-03-12T00:00:00 | [
[
"Lovenia",
"Holy",
""
],
[
"Mahendra",
"Rahmad",
""
],
[
"Akbar",
"Salsabil Maulana",
""
],
[
"Miranda",
"Lester James V.",
""
],
[
"Santoso",
"Jennifer",
""
],
[
"Aco",
"Elyanah",
""
],
[
"Fadhilah",
"Akhdan",
""
],
[
"Mansurov",
"Jonibek",
""
],
[
"Imperial",
"Joseph Marvin",
""
],
[
"Kampman",
"Onno P.",
""
],
[
"Moniz",
"Joel Ruben Antony",
""
],
[
"Habibi",
"Muhammad Ravi Shulthan",
""
],
[
"Hudi",
"Frederikus",
""
],
[
"Montalan",
"Railey",
""
],
[
"Ignatius",
"Ryan",
""
],
[
"Lopo",
"Joanito Agili",
""
],
[
"Nixon",
"William",
""
],
[
"Karlsson",
"Börje F.",
""
],
[
"Jaya",
"James",
""
],
[
"Diandaru",
"Ryandito",
""
],
[
"Gao",
"Yuze",
""
],
[
"Amadeus",
"Patrick",
""
],
[
"Wang",
"Bin",
""
],
[
"Cruz",
"Jan Christian Blaise",
""
],
[
"Whitehouse",
"Chenxi",
""
],
[
"Parmonangan",
"Ivan Halim",
""
],
[
"Khelli",
"Maria",
""
],
[
"Zhang",
"Wenyu",
""
],
[
"Susanto",
"Lucky",
""
],
[
"Ryanda",
"Reynard Adha",
""
],
[
"Hermawan",
"Sonny Lazuardi",
""
],
[
"Velasco",
"Dan John",
""
],
[
"Kautsar",
"Muhammad Dehan Al",
""
],
[
"Hendria",
"Willy Fitra",
""
],
[
"Moslem",
"Yasmin",
""
],
[
"Flynn",
"Noah",
""
],
[
"Adilazuarda",
"Muhammad Farid",
""
],
[
"Li",
"Haochen",
""
],
[
"Lee",
"Johanes",
""
],
[
"Damanhuri",
"R.",
""
],
[
"Sun",
"Shuo",
""
],
[
"Qorib",
"Muhammad Reza",
""
],
[
"Djanibekov",
"Amirbek",
""
],
[
"Leong",
"Wei Qi",
""
],
[
"Do",
"Quyet V.",
""
],
[
"Muennighoff",
"Niklas",
""
],
[
"Pansuwan",
"Tanrada",
""
],
[
"Putra",
"Ilham Firdausi",
""
],
[
"Xu",
"Yan",
""
],
[
"Tai",
"Ngee Chia",
""
],
[
"Purwarianti",
"Ayu",
""
],
[
"Ruder",
"Sebastian",
""
],
[
"Tjhi",
"William",
""
],
[
"Limkonchotiwat",
"Peerat",
""
],
[
"Aji",
"Alham Fikri",
""
],
[
"Keh",
"Sedrick",
""
],
[
"Winata",
"Genta Indra",
""
],
[
"Zhang",
"Ruochen",
""
],
[
"Koto",
"Fajri",
""
],
[
"Yong",
"Zheng-Xin",
""
],
[
"Cahyawijaya",
"Samuel",
""
]
]
| TITLE: SEACrowd: A Multilingual Multimodal Data Hub and Benchmark Suite for
Southeast Asian Languages
ABSTRACT: Southeast Asia (SEA) is a region rich in linguistic diversity and cultural
variety, with over 1,300 indigenous languages and a population of 671 million
people. However, prevailing AI models suffer from a significant lack of
representation of texts, images, and audio datasets from SEA, compromising the
quality of AI models for SEA languages. Evaluating models for SEA languages is
challenging due to the scarcity of high-quality datasets, compounded by the
dominance of English training data, raising concerns about potential cultural
misrepresentation. To address these challenges, we introduce SEACrowd, a
collaborative initiative that consolidates a comprehensive resource hub that
fills the resource gap by providing standardized corpora in nearly 1,000 SEA
languages across three modalities. Through our SEACrowd benchmarks, we assess
the quality of AI models on 36 indigenous languages across 13 tasks, offering
valuable insights into the current AI landscape in SEA. Furthermore, we propose
strategies to facilitate greater AI advancements, maximizing potential utility
and resource equity for the future of AI in SEA.
| no_new_dataset | 0.927888 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.