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
2502.20527
Jake Renzella Dr
Emily Ross, Yuval Kansal, Jake Renzella, Alexandra Vassar, Andrew Taylor
Supervised Fine-Tuning LLMs to Behave as Pedagogical Agents in Programming Education
null
null
null
null
cs.CL cs.CY
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Large language models (LLMs) are increasingly being explored in higher education, yet their effectiveness as teaching agents remains underexamined. In this paper, we present the development of GuideLM, a fine-tuned LLM designed for programming education. GuideLM has been integrated into the Debugging C Compiler (DCC), an educational C compiler that leverages LLMs to generate pedagogically sound error explanations. Previously, DCC relied on off-the-shelf OpenAI models, which, while accurate, often over-assisted students by directly providing solutions despite contrary prompting. To address this, we employed supervised fine-tuning (SFT) on a dataset of 528 student-question/teacher-answer pairs, creating two models: GuideLM and GuideLM-mini, fine-tuned on ChatGPT-4o and 4o-mini, respectively. We conducted an expert analysis of 400 responses per model, comparing their pedagogical effectiveness against base OpenAI models. Our evaluation, grounded in constructivism and cognitive load theory, assessed factors such as conceptual scaffolding, clarity, and Socratic guidance. Results indicate that GuideLM and GuideLM-mini improve pedagogical performance, with an 8% increase in Socratic guidance and a 58% improvement in economy of words compared to GPT-4o. However, this refinement comes at the cost of a slight reduction in general accuracy. While further work is needed, our findings suggest that fine-tuning LLMs with targeted datasets is a promising approach for developing models better suited to educational contexts.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 21:23:56 GMT" } ]
2025-03-03T00:00:00
[ [ "Ross", "Emily", "" ], [ "Kansal", "Yuval", "" ], [ "Renzella", "Jake", "" ], [ "Vassar", "Alexandra", "" ], [ "Taylor", "Andrew", "" ] ]
TITLE: Supervised Fine-Tuning LLMs to Behave as Pedagogical Agents in Programming Education ABSTRACT: Large language models (LLMs) are increasingly being explored in higher education, yet their effectiveness as teaching agents remains underexamined. In this paper, we present the development of GuideLM, a fine-tuned LLM designed for programming education. GuideLM has been integrated into the Debugging C Compiler (DCC), an educational C compiler that leverages LLMs to generate pedagogically sound error explanations. Previously, DCC relied on off-the-shelf OpenAI models, which, while accurate, often over-assisted students by directly providing solutions despite contrary prompting. To address this, we employed supervised fine-tuning (SFT) on a dataset of 528 student-question/teacher-answer pairs, creating two models: GuideLM and GuideLM-mini, fine-tuned on ChatGPT-4o and 4o-mini, respectively. We conducted an expert analysis of 400 responses per model, comparing their pedagogical effectiveness against base OpenAI models. Our evaluation, grounded in constructivism and cognitive load theory, assessed factors such as conceptual scaffolding, clarity, and Socratic guidance. Results indicate that GuideLM and GuideLM-mini improve pedagogical performance, with an 8% increase in Socratic guidance and a 58% improvement in economy of words compared to GPT-4o. However, this refinement comes at the cost of a slight reduction in general accuracy. While further work is needed, our findings suggest that fine-tuning LLMs with targeted datasets is a promising approach for developing models better suited to educational contexts.
no_new_dataset
0.943815
2502.20537
Phil\'emon Houdaille
Phil\'emon Houdaille (University of Rennes, France / CNRS, France / Inria, France), Djamel Eddine Khelladi (CNRS, France / University of Rennes, France), Benoit Combemale (University of Rennes, France), Gunter Mussbacher (McGill University, Canada / Inria, France), Tijs van der Storm (CWI, Netherlands / University of Groningen, Netherlands)
PolyDebug: A Framework for Polyglot Debugging
null
The Art, Science, and Engineering of Programming, 2025, Vol. 10, Issue 1, Article 13
10.22152/programming-journal.org/2026/10/13
null
cs.PL
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
As software grows increasingly complex, the quantity and diversity of concerns to be addressed also rises. To answer this diversity of concerns, developers may end up using multiple programming languages in a single software project, a practice known as polyglot programming. This practice has gained momentum with the rise of execution platforms capable of supporting polyglot systems. However, despite this momentum, there is a notable lack of development tooling support for developers working on polyglot programs, such as in debugging facilities. Not all polyglot execution platforms provide debugging capabilities, and for those that do, implementing support for new languages can be costly. This paper addresses this gap by introducing a novel debugger framework that is language-agnostic yet leverages existing language-specific debuggers. The proposed framework is dynamically extensible to accommodate the evolving combination of languages used in polyglot software development. It utilizes the Debug Adapter Protocol (DAP) to integrate and coordinate existing debuggers within a debugging session. We found that using our approach, we were able to implement polyglot debugging support for three different languages with little development effort. We also found that our debugger did not introduce an overhead significant enough to hinder debugging tasks in many scenarios; however performance did deteriorate with the amount of polyglot calls, making the approach not suitable for every polyglot program structure. The effectiveness of this approach is demonstrated through the development of a prototype, PolyDebug, and its application to use cases involving C, JavaScript, and Python. We evaluated PolyDebug on a dataset of traditional benchmark programs, modified to fit our criteria of polyglot programs. We also assessed the development effort by measuring the source lines of code (SLOC) for the prototype as a whole as well as its components. Debugging is a fundamental part of developing and maintaining software. Lack of debug tools can lead to difficulty in locating software bugs and slow down the development process. We believe this work is relevant to help provide developers proper debugging support regardless of the runtime environment.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 21:39:05 GMT" } ]
2025-03-03T00:00:00
[ [ "Houdaille", "Philémon", "", "University of Rennes, France / CNRS, France /\n Inria, France" ], [ "Khelladi", "Djamel Eddine", "", "CNRS, France / University of Rennes,\n France" ], [ "Combemale", "Benoit", "", "University of Rennes, France" ], [ "Mussbacher", "Gunter", "", "McGill University, Canada / Inria, France" ], [ "van der Storm", "Tijs", "", "CWI,\n Netherlands / University of Groningen, Netherlands" ] ]
TITLE: PolyDebug: A Framework for Polyglot Debugging ABSTRACT: As software grows increasingly complex, the quantity and diversity of concerns to be addressed also rises. To answer this diversity of concerns, developers may end up using multiple programming languages in a single software project, a practice known as polyglot programming. This practice has gained momentum with the rise of execution platforms capable of supporting polyglot systems. However, despite this momentum, there is a notable lack of development tooling support for developers working on polyglot programs, such as in debugging facilities. Not all polyglot execution platforms provide debugging capabilities, and for those that do, implementing support for new languages can be costly. This paper addresses this gap by introducing a novel debugger framework that is language-agnostic yet leverages existing language-specific debuggers. The proposed framework is dynamically extensible to accommodate the evolving combination of languages used in polyglot software development. It utilizes the Debug Adapter Protocol (DAP) to integrate and coordinate existing debuggers within a debugging session. We found that using our approach, we were able to implement polyglot debugging support for three different languages with little development effort. We also found that our debugger did not introduce an overhead significant enough to hinder debugging tasks in many scenarios; however performance did deteriorate with the amount of polyglot calls, making the approach not suitable for every polyglot program structure. The effectiveness of this approach is demonstrated through the development of a prototype, PolyDebug, and its application to use cases involving C, JavaScript, and Python. We evaluated PolyDebug on a dataset of traditional benchmark programs, modified to fit our criteria of polyglot programs. We also assessed the development effort by measuring the source lines of code (SLOC) for the prototype as a whole as well as its components. Debugging is a fundamental part of developing and maintaining software. Lack of debug tools can lead to difficulty in locating software bugs and slow down the development process. We believe this work is relevant to help provide developers proper debugging support regardless of the runtime environment.
no_new_dataset
0.923454
2502.20545
Shiwei Liu
Kechen Li, Wenqi Zhu, Coralia Cartis, Tianbo Ji, Shiwei Liu
SoS1: O1 and R1-Like Reasoning LLMs are Sum-of-Square Solvers
null
null
null
null
cs.LG
http://creativecommons.org/licenses/by/4.0/
Large Language Models (LLMs) have achieved human-level proficiency across diverse tasks, but their ability to perform rigorous mathematical problem solving remains an open challenge. In this work, we investigate a fundamental yet computationally intractable problem: determining whether a given multivariate polynomial is nonnegative. This problem, closely related to Hilbert's Seventeenth Problem, plays a crucial role in global polynomial optimization and has applications in various fields. First, we introduce SoS-1K, a meticulously curated dataset of approximately 1,000 polynomials, along with expert-designed reasoning instructions based on five progressively challenging criteria. Evaluating multiple state-of-the-art LLMs, we find that without structured guidance, all models perform only slightly above the random guess baseline 50%. However, high-quality reasoning instructions significantly improve accuracy, boosting performance up to 81%. Furthermore, our 7B model, SoS-7B, fine-tuned on SoS-1K for just 4 hours, outperforms the 671B DeepSeek-V3 and GPT-4o-mini in accuracy while only requiring 1.8% and 5% of the computation time needed for letters, respectively. Our findings highlight the potential of LLMs to push the boundaries of mathematical reasoning and tackle NP-hard problems.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 21:41:43 GMT" } ]
2025-03-03T00:00:00
[ [ "Li", "Kechen", "" ], [ "Zhu", "Wenqi", "" ], [ "Cartis", "Coralia", "" ], [ "Ji", "Tianbo", "" ], [ "Liu", "Shiwei", "" ] ]
TITLE: SoS1: O1 and R1-Like Reasoning LLMs are Sum-of-Square Solvers ABSTRACT: Large Language Models (LLMs) have achieved human-level proficiency across diverse tasks, but their ability to perform rigorous mathematical problem solving remains an open challenge. In this work, we investigate a fundamental yet computationally intractable problem: determining whether a given multivariate polynomial is nonnegative. This problem, closely related to Hilbert's Seventeenth Problem, plays a crucial role in global polynomial optimization and has applications in various fields. First, we introduce SoS-1K, a meticulously curated dataset of approximately 1,000 polynomials, along with expert-designed reasoning instructions based on five progressively challenging criteria. Evaluating multiple state-of-the-art LLMs, we find that without structured guidance, all models perform only slightly above the random guess baseline 50%. However, high-quality reasoning instructions significantly improve accuracy, boosting performance up to 81%. Furthermore, our 7B model, SoS-7B, fine-tuned on SoS-1K for just 4 hours, outperforms the 671B DeepSeek-V3 and GPT-4o-mini in accuracy while only requiring 1.8% and 5% of the computation time needed for letters, respectively. Our findings highlight the potential of LLMs to push the boundaries of mathematical reasoning and tackle NP-hard problems.
new_dataset
0.958693
2502.20548
Jin Peng Zhou
Jin Peng Zhou, Kaiwen Wang, Jonathan Chang, Zhaolin Gao, Nathan Kallus, Kilian Q. Weinberger, Kiant\'e Brantley, Wen Sun
$Q\sharp$: Provably Optimal Distributional RL for LLM Post-Training
null
null
null
null
cs.LG cs.AI cs.CL
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Reinforcement learning (RL) post-training is crucial for LLM alignment and reasoning, but existing policy-based methods, such as PPO and DPO, can fall short of fixing shortcuts inherited from pre-training. In this work, we introduce $Q\sharp$, a value-based algorithm for KL-regularized RL that guides the reference policy using the optimal regularized $Q$ function. We propose to learn the optimal $Q$ function using distributional RL on an aggregated online dataset. Unlike prior value-based baselines that guide the model using unregularized $Q$-values, our method is theoretically principled and provably learns the optimal policy for the KL-regularized RL problem. Empirically, $Q\sharp$ outperforms prior baselines in math reasoning benchmarks while maintaining a smaller KL divergence to the reference policy. Theoretically, we establish a reduction from KL-regularized RL to no-regret online learning, providing the first bounds for deterministic MDPs under only realizability. Thanks to distributional RL, our bounds are also variance-dependent and converge faster when the reference policy has small variance. In sum, our results highlight $Q\sharp$ as an effective approach for post-training LLMs, offering both improved performance and theoretical guarantees. The code can be found at https://github.com/jinpz/q_sharp.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 21:43:00 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhou", "Jin Peng", "" ], [ "Wang", "Kaiwen", "" ], [ "Chang", "Jonathan", "" ], [ "Gao", "Zhaolin", "" ], [ "Kallus", "Nathan", "" ], [ "Weinberger", "Kilian Q.", "" ], [ "Brantley", "Kianté", "" ], [ "Sun", "Wen", "" ] ]
TITLE: $Q\sharp$: Provably Optimal Distributional RL for LLM Post-Training ABSTRACT: Reinforcement learning (RL) post-training is crucial for LLM alignment and reasoning, but existing policy-based methods, such as PPO and DPO, can fall short of fixing shortcuts inherited from pre-training. In this work, we introduce $Q\sharp$, a value-based algorithm for KL-regularized RL that guides the reference policy using the optimal regularized $Q$ function. We propose to learn the optimal $Q$ function using distributional RL on an aggregated online dataset. Unlike prior value-based baselines that guide the model using unregularized $Q$-values, our method is theoretically principled and provably learns the optimal policy for the KL-regularized RL problem. Empirically, $Q\sharp$ outperforms prior baselines in math reasoning benchmarks while maintaining a smaller KL divergence to the reference policy. Theoretically, we establish a reduction from KL-regularized RL to no-regret online learning, providing the first bounds for deterministic MDPs under only realizability. Thanks to distributional RL, our bounds are also variance-dependent and converge faster when the reference policy has small variance. In sum, our results highlight $Q\sharp$ as an effective approach for post-training LLMs, offering both improved performance and theoretical guarantees. The code can be found at https://github.com/jinpz/q_sharp.
no_new_dataset
0.942082
2502.20552
Botond Barta
Botond Barta, Endre Hamerlik, Mil\'an Konor Nyist, Judit \'Acs
HuAMR: A Hungarian AMR Parser and Dataset
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
We present HuAMR, the first Abstract Meaning Representation (AMR) dataset and a suite of large language model-based AMR parsers for Hungarian, targeting the scarcity of semantic resources for non-English languages. To create HuAMR, we employed Llama-3.1-70B to automatically generate silver-standard AMR annotations, which we then refined manually to ensure quality. Building on this dataset, we investigate how different model architectures - mT5 Large and Llama-3.2-1B - and fine-tuning strategies affect AMR parsing performance. While incorporating silver-standard AMRs from Llama-3.1-70B into the training data of smaller models does not consistently boost overall scores, our results show that these techniques effectively enhance parsing accuracy on Hungarian news data (the domain of HuAMR). We evaluate our parsers using Smatch scores and confirm the potential of HuAMR and our parsers for advancing semantic parsing research.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 21:48:11 GMT" } ]
2025-03-03T00:00:00
[ [ "Barta", "Botond", "" ], [ "Hamerlik", "Endre", "" ], [ "Nyist", "Milán Konor", "" ], [ "Ács", "Judit", "" ] ]
TITLE: HuAMR: A Hungarian AMR Parser and Dataset ABSTRACT: We present HuAMR, the first Abstract Meaning Representation (AMR) dataset and a suite of large language model-based AMR parsers for Hungarian, targeting the scarcity of semantic resources for non-English languages. To create HuAMR, we employed Llama-3.1-70B to automatically generate silver-standard AMR annotations, which we then refined manually to ensure quality. Building on this dataset, we investigate how different model architectures - mT5 Large and Llama-3.2-1B - and fine-tuning strategies affect AMR parsing performance. While incorporating silver-standard AMRs from Llama-3.1-70B into the training data of smaller models does not consistently boost overall scores, our results show that these techniques effectively enhance parsing accuracy on Hungarian news data (the domain of HuAMR). We evaluate our parsers using Smatch scores and confirm the potential of HuAMR and our parsers for advancing semantic parsing research.
new_dataset
0.964623
2502.20562
Joana C. Costa
Joana C. Costa and Tiago Roxo and Hugo Proen\c{c}a and Pedro R. M. In\'acio
LISArD: Learning Image Similarity to Defend Against Gray-box Adversarial Attacks
null
null
null
null
cs.CV cs.CR cs.LG
http://creativecommons.org/licenses/by-nc-sa/4.0/
State-of-the-art defense mechanisms are typically evaluated in the context of white-box attacks, which is not realistic, as it assumes the attacker can access the gradients of the target network. To protect against this scenario, Adversarial Training (AT) and Adversarial Distillation (AD) include adversarial examples during the training phase, and Adversarial Purification uses a generative model to reconstruct all the images given to the classifier. This paper considers an even more realistic evaluation scenario: gray-box attacks, which assume that the attacker knows the architecture and the dataset used to train the target network, but cannot access its gradients. We provide empirical evidence that models are vulnerable to gray-box attacks and propose LISArD, a defense mechanism that does not increase computational and temporal costs but provides robustness against gray-box and white-box attacks without including AT. Our method approximates a cross-correlation matrix, created with the embeddings of perturbed and clean images, to a diagonal matrix while simultaneously conducting classification learning. Our results show that LISArD can effectively protect against gray-box attacks, can be used in multiple architectures, and carries over its resilience to the white-box scenario. Also, state-of-the-art AD models underperform greatly when removing AT and/or moving to gray-box settings, highlighting the lack of robustness from existing approaches to perform in various conditions (aside from white-box settings). All the source code is available at https://github.com/Joana-Cabral/LISArD.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 22:02:06 GMT" } ]
2025-03-03T00:00:00
[ [ "Costa", "Joana C.", "" ], [ "Roxo", "Tiago", "" ], [ "Proença", "Hugo", "" ], [ "Inácio", "Pedro R. M.", "" ] ]
TITLE: LISArD: Learning Image Similarity to Defend Against Gray-box Adversarial Attacks ABSTRACT: State-of-the-art defense mechanisms are typically evaluated in the context of white-box attacks, which is not realistic, as it assumes the attacker can access the gradients of the target network. To protect against this scenario, Adversarial Training (AT) and Adversarial Distillation (AD) include adversarial examples during the training phase, and Adversarial Purification uses a generative model to reconstruct all the images given to the classifier. This paper considers an even more realistic evaluation scenario: gray-box attacks, which assume that the attacker knows the architecture and the dataset used to train the target network, but cannot access its gradients. We provide empirical evidence that models are vulnerable to gray-box attacks and propose LISArD, a defense mechanism that does not increase computational and temporal costs but provides robustness against gray-box and white-box attacks without including AT. Our method approximates a cross-correlation matrix, created with the embeddings of perturbed and clean images, to a diagonal matrix while simultaneously conducting classification learning. Our results show that LISArD can effectively protect against gray-box attacks, can be used in multiple architectures, and carries over its resilience to the white-box scenario. Also, state-of-the-art AD models underperform greatly when removing AT and/or moving to gray-box settings, highlighting the lack of robustness from existing approaches to perform in various conditions (aside from white-box settings). All the source code is available at https://github.com/Joana-Cabral/LISArD.
no_new_dataset
0.941975
2502.20571
David Anastasiu
Yanhong Li and David C. Anastasiu
PFformer: A Position-Free Transformer Variant for Extreme-Adaptive Multivariate Time Series Forecasting
PAKDD 2025 special session on Data Science: Foundations and Applications (DSFA)
null
null
null
cs.LG cs.AI
http://creativecommons.org/licenses/by/4.0/
Multivariate time series (MTS) forecasting is vital in fields like weather, energy, and finance. However, despite deep learning advancements, traditional Transformer-based models often diminish the effect of crucial inter-variable relationships by singular token embedding and struggle to effectively capture complex dependencies among variables, especially in datasets with rare or extreme events. These events create significant imbalances and lead to high skewness, complicating accurate prediction efforts. This study introduces PFformer, a position-free Transformer-based model designed for single-target MTS forecasting, specifically for challenging datasets characterized by extreme variability. PFformer integrates two novel embedding strategies: Enhanced Feature-based Embedding (EFE) and Auto-Encoder-based Embedding (AEE). EFE effectively encodes inter-variable dependencies by mapping related sequence subsets to high-dimensional spaces without positional constraints, enhancing the encoder's functionality. PFformer shows superior forecasting accuracy without the traditional limitations of positional encoding in MTS modeling. We evaluated PFformer across four challenging datasets, focusing on two key forecasting scenarios: long sequence prediction for 3 days ahead and rolling predictions every four hours to reflect real-time decision-making processes in water management. PFformer demonstrated remarkable improvements, from 20% to 60%, compared with state-of-the-art models.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 22:21:27 GMT" } ]
2025-03-03T00:00:00
[ [ "Li", "Yanhong", "" ], [ "Anastasiu", "David C.", "" ] ]
TITLE: PFformer: A Position-Free Transformer Variant for Extreme-Adaptive Multivariate Time Series Forecasting ABSTRACT: Multivariate time series (MTS) forecasting is vital in fields like weather, energy, and finance. However, despite deep learning advancements, traditional Transformer-based models often diminish the effect of crucial inter-variable relationships by singular token embedding and struggle to effectively capture complex dependencies among variables, especially in datasets with rare or extreme events. These events create significant imbalances and lead to high skewness, complicating accurate prediction efforts. This study introduces PFformer, a position-free Transformer-based model designed for single-target MTS forecasting, specifically for challenging datasets characterized by extreme variability. PFformer integrates two novel embedding strategies: Enhanced Feature-based Embedding (EFE) and Auto-Encoder-based Embedding (AEE). EFE effectively encodes inter-variable dependencies by mapping related sequence subsets to high-dimensional spaces without positional constraints, enhancing the encoder's functionality. PFformer shows superior forecasting accuracy without the traditional limitations of positional encoding in MTS modeling. We evaluated PFformer across four challenging datasets, focusing on two key forecasting scenarios: long sequence prediction for 3 days ahead and rolling predictions every four hours to reflect real-time decision-making processes in water management. PFformer demonstrated remarkable improvements, from 20% to 60%, compared with state-of-the-art models.
no_new_dataset
0.946695
2502.20572
Huthaifa I. Ashqar
Mohammad Abu Tami, Mohammed Elhenawy, and Huthaifa I. Ashqar
HazardNet: A Small-Scale Vision Language Model for Real-Time Traffic Safety Detection at Edge Devices
null
null
null
null
cs.CV cs.CL
http://creativecommons.org/licenses/by/4.0/
Traffic safety remains a vital concern in contemporary urban settings, intensified by the increase of vehicles and the complicated nature of road networks. Traditional safety-critical event detection systems predominantly rely on sensor-based approaches and conventional machine learning algorithms, necessitating extensive data collection and complex training processes to adhere to traffic safety regulations. This paper introduces HazardNet, a small-scale Vision Language Model designed to enhance traffic safety by leveraging the reasoning capabilities of advanced language and vision models. We built HazardNet by fine-tuning the pre-trained Qwen2-VL-2B model, chosen for its superior performance among open-source alternatives and its compact size of two billion parameters. This helps to facilitate deployment on edge devices with efficient inference throughput. In addition, we present HazardQA, a novel Vision Question Answering (VQA) dataset constructed specifically for training HazardNet on real-world scenarios involving safety-critical events. Our experimental results show that the fine-tuned HazardNet outperformed the base model up to an 89% improvement in F1-Score and has comparable results with improvement in some cases reach up to 6% when compared to larger models, such as GPT-4o. These advancements underscore the potential of HazardNet in providing real-time, reliable traffic safety event detection, thereby contributing to reduced accidents and improved traffic management in urban environments. Both HazardNet model and the HazardQA dataset are available at https://huggingface.co/Tami3/HazardNet and https://huggingface.co/datasets/Tami3/HazardQA, respectively.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 22:21:45 GMT" } ]
2025-03-03T00:00:00
[ [ "Tami", "Mohammad Abu", "" ], [ "Elhenawy", "Mohammed", "" ], [ "Ashqar", "Huthaifa I.", "" ] ]
TITLE: HazardNet: A Small-Scale Vision Language Model for Real-Time Traffic Safety Detection at Edge Devices ABSTRACT: Traffic safety remains a vital concern in contemporary urban settings, intensified by the increase of vehicles and the complicated nature of road networks. Traditional safety-critical event detection systems predominantly rely on sensor-based approaches and conventional machine learning algorithms, necessitating extensive data collection and complex training processes to adhere to traffic safety regulations. This paper introduces HazardNet, a small-scale Vision Language Model designed to enhance traffic safety by leveraging the reasoning capabilities of advanced language and vision models. We built HazardNet by fine-tuning the pre-trained Qwen2-VL-2B model, chosen for its superior performance among open-source alternatives and its compact size of two billion parameters. This helps to facilitate deployment on edge devices with efficient inference throughput. In addition, we present HazardQA, a novel Vision Question Answering (VQA) dataset constructed specifically for training HazardNet on real-world scenarios involving safety-critical events. Our experimental results show that the fine-tuned HazardNet outperformed the base model up to an 89% improvement in F1-Score and has comparable results with improvement in some cases reach up to 6% when compared to larger models, such as GPT-4o. These advancements underscore the potential of HazardNet in providing real-time, reliable traffic safety event detection, thereby contributing to reduced accidents and improved traffic management in urban environments. Both HazardNet model and the HazardQA dataset are available at https://huggingface.co/Tami3/HazardNet and https://huggingface.co/datasets/Tami3/HazardQA, respectively.
new_dataset
0.965152
2502.20582
Zihao He
Javin Liu, Aryan Vats, Zihao He
CS-PaperSum: A Large-Scale Dataset of AI-Generated Summaries for Scientific Papers
null
null
null
null
cs.IR
http://creativecommons.org/licenses/by/4.0/
The rapid expansion of scientific literature in computer science presents challenges in tracking research trends and extracting key insights. Existing datasets provide metadata but lack structured summaries that capture core contributions and methodologies. We introduce CS-PaperSum, a large-scale dataset of 91,919 papers from 31 top-tier computer science conferences, enriched with AI-generated structured summaries using ChatGPT. To assess summary quality, we conduct embedding alignment analysis and keyword overlap analysis, demonstrating strong preservation of key concepts. We further present a case study on AI research trends, highlighting shifts in methodologies and interdisciplinary crossovers, including the rise of self-supervised learning, retrieval-augmented generation, and multimodal AI. Our dataset enables automated literature analysis, research trend forecasting, and AI-driven scientific discovery, providing a valuable resource for researchers, policymakers, and scientific information retrieval systems.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 22:48:35 GMT" } ]
2025-03-03T00:00:00
[ [ "Liu", "Javin", "" ], [ "Vats", "Aryan", "" ], [ "He", "Zihao", "" ] ]
TITLE: CS-PaperSum: A Large-Scale Dataset of AI-Generated Summaries for Scientific Papers ABSTRACT: The rapid expansion of scientific literature in computer science presents challenges in tracking research trends and extracting key insights. Existing datasets provide metadata but lack structured summaries that capture core contributions and methodologies. We introduce CS-PaperSum, a large-scale dataset of 91,919 papers from 31 top-tier computer science conferences, enriched with AI-generated structured summaries using ChatGPT. To assess summary quality, we conduct embedding alignment analysis and keyword overlap analysis, demonstrating strong preservation of key concepts. We further present a case study on AI research trends, highlighting shifts in methodologies and interdisciplinary crossovers, including the rise of self-supervised learning, retrieval-augmented generation, and multimodal AI. Our dataset enables automated literature analysis, research trend forecasting, and AI-driven scientific discovery, providing a valuable resource for researchers, policymakers, and scientific information retrieval systems.
new_dataset
0.958963
2502.20583
Keisuke Kamahori
Keisuke Kamahori, Jungo Kasai, Noriyuki Kojima, Baris Kasikci
LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation
null
null
null
null
cs.LG cs.AI cs.SD eess.AS
http://creativecommons.org/licenses/by/4.0/
Modern automatic speech recognition (ASR) models, such as OpenAI's Whisper, rely on deep encoder-decoder architectures, and their encoders are a critical bottleneck for efficient deployment due to high computational intensity. We introduce LiteASR, a low-rank compression scheme for ASR encoders that significantly reduces inference costs while maintaining transcription accuracy. Our approach leverages the strong low-rank properties observed in intermediate activations: by applying principal component analysis (PCA) with a small calibration dataset, we approximate linear transformations with a chain of low-rank matrix multiplications, and further optimize self-attention to work in the reduced dimension. Evaluation results show that our method can compress Whisper large-v3's encoder size by over 50%, matching Whisper medium's size with better transcription accuracy, thereby establishing a new Pareto-optimal frontier of efficiency and performance. The code of LiteASR is available at https://github.com/efeslab/LiteASR.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 22:52:21 GMT" } ]
2025-03-03T00:00:00
[ [ "Kamahori", "Keisuke", "" ], [ "Kasai", "Jungo", "" ], [ "Kojima", "Noriyuki", "" ], [ "Kasikci", "Baris", "" ] ]
TITLE: LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation ABSTRACT: Modern automatic speech recognition (ASR) models, such as OpenAI's Whisper, rely on deep encoder-decoder architectures, and their encoders are a critical bottleneck for efficient deployment due to high computational intensity. We introduce LiteASR, a low-rank compression scheme for ASR encoders that significantly reduces inference costs while maintaining transcription accuracy. Our approach leverages the strong low-rank properties observed in intermediate activations: by applying principal component analysis (PCA) with a small calibration dataset, we approximate linear transformations with a chain of low-rank matrix multiplications, and further optimize self-attention to work in the reduced dimension. Evaluation results show that our method can compress Whisper large-v3's encoder size by over 50%, matching Whisper medium's size with better transcription accuracy, thereby establishing a new Pareto-optimal frontier of efficiency and performance. The code of LiteASR is available at https://github.com/efeslab/LiteASR.
no_new_dataset
0.94428
2502.20596
Linh Ngo
Nguyen Xuan Thanh, Anh Duc Le, Quyen Tran, Thanh-Thien Le, Linh Ngo Van, Thien Huu Nguyen
Few-Shot, No Problem: Descriptive Continual Relation Extraction
Accepted to AAAI 2025
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
Few-shot Continual Relation Extraction is a crucial challenge for enabling AI systems to identify and adapt to evolving relationships in dynamic real-world domains. Traditional memory-based approaches often overfit to limited samples, failing to reinforce old knowledge, with the scarcity of data in few-shot scenarios further exacerbating these issues by hindering effective data augmentation in the latent space. In this paper, we propose a novel retrieval-based solution, starting with a large language model to generate descriptions for each relation. From these descriptions, we introduce a bi-encoder retrieval training paradigm to enrich both sample and class representation learning. Leveraging these enhanced representations, we design a retrieval-based prediction method where each sample "retrieves" the best fitting relation via a reciprocal rank fusion score that integrates both relation description vectors and class prototypes. Extensive experiments on multiple datasets demonstrate that our method significantly advances the state-of-the-art by maintaining robust performance across sequential tasks, effectively addressing catastrophic forgetting.
[ { "version": "v1", "created": "Thu, 27 Feb 2025 23:44:30 GMT" } ]
2025-03-03T00:00:00
[ [ "Thanh", "Nguyen Xuan", "" ], [ "Le", "Anh Duc", "" ], [ "Tran", "Quyen", "" ], [ "Le", "Thanh-Thien", "" ], [ "Van", "Linh Ngo", "" ], [ "Nguyen", "Thien Huu", "" ] ]
TITLE: Few-Shot, No Problem: Descriptive Continual Relation Extraction ABSTRACT: Few-shot Continual Relation Extraction is a crucial challenge for enabling AI systems to identify and adapt to evolving relationships in dynamic real-world domains. Traditional memory-based approaches often overfit to limited samples, failing to reinforce old knowledge, with the scarcity of data in few-shot scenarios further exacerbating these issues by hindering effective data augmentation in the latent space. In this paper, we propose a novel retrieval-based solution, starting with a large language model to generate descriptions for each relation. From these descriptions, we introduce a bi-encoder retrieval training paradigm to enrich both sample and class representation learning. Leveraging these enhanced representations, we design a retrieval-based prediction method where each sample "retrieves" the best fitting relation via a reciprocal rank fusion score that integrates both relation description vectors and class prototypes. Extensive experiments on multiple datasets demonstrate that our method significantly advances the state-of-the-art by maintaining robust performance across sequential tasks, effectively addressing catastrophic forgetting.
no_new_dataset
0.947962
2502.20603
Ira Shokar Mr
Ira J. S. Shokar, Peter H. Haynes, Rich R. Kerswell
Deep Learning of the Evolution Operator Enables Forecasting of Out-of-Training Dynamics in Chaotic Systems
null
null
null
null
cs.LG math.DS nlin.CD
http://creativecommons.org/licenses/by/4.0/
We demonstrate that a deep learning emulator for chaotic systems can forecast phenomena absent from training data. Using the Kuramoto-Sivashinsky and beta-plane turbulence models, we evaluate the emulator through scenarios probing the fundamental phenomena of both systems: forecasting spontaneous relaminarisation, capturing initialisation of arbitrary chaotic states, zero-shot prediction of dynamics with parameter values outside of the training range, and characterisation of dynamical statistics from artificially restricted training datasets. Our results show that deep learning emulators can uncover emergent behaviours and rare events in complex systems by learning underlying mathematical rules, rather than merely mimicking observed patterns.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:07:18 GMT" } ]
2025-03-03T00:00:00
[ [ "Shokar", "Ira J. S.", "" ], [ "Haynes", "Peter H.", "" ], [ "Kerswell", "Rich R.", "" ] ]
TITLE: Deep Learning of the Evolution Operator Enables Forecasting of Out-of-Training Dynamics in Chaotic Systems ABSTRACT: We demonstrate that a deep learning emulator for chaotic systems can forecast phenomena absent from training data. Using the Kuramoto-Sivashinsky and beta-plane turbulence models, we evaluate the emulator through scenarios probing the fundamental phenomena of both systems: forecasting spontaneous relaminarisation, capturing initialisation of arbitrary chaotic states, zero-shot prediction of dynamics with parameter values outside of the training range, and characterisation of dynamical statistics from artificially restricted training datasets. Our results show that deep learning emulators can uncover emergent behaviours and rare events in complex systems by learning underlying mathematical rules, rather than merely mimicking observed patterns.
no_new_dataset
0.949389
2502.20604
Hao Xuan
Hao Xuan, Bokai Yang, Xingyu Li
Exploring the Impact of Temperature Scaling in Softmax for Classification and Adversarial Robustness
null
null
null
null
cs.LG cs.AI
http://creativecommons.org/licenses/by/4.0/
The softmax function is a fundamental component in deep learning. This study delves into the often-overlooked parameter within the softmax function, known as "temperature," providing novel insights into the practical and theoretical aspects of temperature scaling for image classification. Our empirical studies, adopting convolutional neural networks and transformers on multiple benchmark datasets, reveal that moderate temperatures generally introduce better overall performance. Through extensive experiments and rigorous theoretical analysis, we explore the role of temperature scaling in model training and unveil that temperature not only influences learning step size but also shapes the model's optimization direction. Moreover, for the first time, we discover a surprising benefit of elevated temperatures: enhanced model robustness against common corruption, natural perturbation, and non-targeted adversarial attacks like Projected Gradient Descent. We extend our discoveries to adversarial training, demonstrating that, compared to the standard softmax function with the default temperature value, higher temperatures have the potential to enhance adversarial training. The insights of this work open new avenues for improving model performance and security in deep learning applications.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:07:45 GMT" } ]
2025-03-03T00:00:00
[ [ "Xuan", "Hao", "" ], [ "Yang", "Bokai", "" ], [ "Li", "Xingyu", "" ] ]
TITLE: Exploring the Impact of Temperature Scaling in Softmax for Classification and Adversarial Robustness ABSTRACT: The softmax function is a fundamental component in deep learning. This study delves into the often-overlooked parameter within the softmax function, known as "temperature," providing novel insights into the practical and theoretical aspects of temperature scaling for image classification. Our empirical studies, adopting convolutional neural networks and transformers on multiple benchmark datasets, reveal that moderate temperatures generally introduce better overall performance. Through extensive experiments and rigorous theoretical analysis, we explore the role of temperature scaling in model training and unveil that temperature not only influences learning step size but also shapes the model's optimization direction. Moreover, for the first time, we discover a surprising benefit of elevated temperatures: enhanced model robustness against common corruption, natural perturbation, and non-targeted adversarial attacks like Projected Gradient Descent. We extend our discoveries to adversarial training, demonstrating that, compared to the standard softmax function with the default temperature value, higher temperatures have the potential to enhance adversarial training. The insights of this work open new avenues for improving model performance and security in deep learning applications.
no_new_dataset
0.946646
2502.20607
Zhefan Xu
Zhefan Xu, Haoyu Shen, Xinming Han, Hanyu Jin, Kanlong Ye, Kenji Shimada
LV-DOT: LiDAR-visual dynamic obstacle detection and tracking for autonomous robot navigation
8 pages, 7 figures, 2 tables
null
null
null
cs.RO
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Accurate perception of dynamic obstacles is essential for autonomous robot navigation in indoor environments. Although sophisticated 3D object detection and tracking methods have been investigated and developed thoroughly in the fields of computer vision and autonomous driving, their demands on expensive and high-accuracy sensor setups and substantial computational resources from large neural networks make them unsuitable for indoor robotics. Recently, more lightweight perception algorithms leveraging onboard cameras or LiDAR sensors have emerged as promising alternatives. However, relying on a single sensor poses significant limitations: cameras have limited fields of view and can suffer from high noise, whereas LiDAR sensors operate at lower frequencies and lack the richness of visual features. To address this limitation, we propose a dynamic obstacle detection and tracking framework that uses both onboard camera and LiDAR data to enable lightweight and accurate perception. Our proposed method expands on our previous ensemble detection approach, which integrates outputs from multiple low-accuracy but computationally efficient detectors to ensure real-time performance on the onboard computer. In this work, we propose a more robust fusion strategy that integrates both LiDAR and visual data to enhance detection accuracy further. We then utilize a tracking module that adopts feature-based object association and the Kalman filter to track and estimate detected obstacles' states. Besides, a dynamic obstacle classification algorithm is designed to robustly identify moving objects. The dataset evaluation demonstrates a better perception performance compared to benchmark methods. The physical experiments on a quadcopter robot confirms the feasibility for real-world navigation.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:12:35 GMT" } ]
2025-03-03T00:00:00
[ [ "Xu", "Zhefan", "" ], [ "Shen", "Haoyu", "" ], [ "Han", "Xinming", "" ], [ "Jin", "Hanyu", "" ], [ "Ye", "Kanlong", "" ], [ "Shimada", "Kenji", "" ] ]
TITLE: LV-DOT: LiDAR-visual dynamic obstacle detection and tracking for autonomous robot navigation ABSTRACT: Accurate perception of dynamic obstacles is essential for autonomous robot navigation in indoor environments. Although sophisticated 3D object detection and tracking methods have been investigated and developed thoroughly in the fields of computer vision and autonomous driving, their demands on expensive and high-accuracy sensor setups and substantial computational resources from large neural networks make them unsuitable for indoor robotics. Recently, more lightweight perception algorithms leveraging onboard cameras or LiDAR sensors have emerged as promising alternatives. However, relying on a single sensor poses significant limitations: cameras have limited fields of view and can suffer from high noise, whereas LiDAR sensors operate at lower frequencies and lack the richness of visual features. To address this limitation, we propose a dynamic obstacle detection and tracking framework that uses both onboard camera and LiDAR data to enable lightweight and accurate perception. Our proposed method expands on our previous ensemble detection approach, which integrates outputs from multiple low-accuracy but computationally efficient detectors to ensure real-time performance on the onboard computer. In this work, we propose a more robust fusion strategy that integrates both LiDAR and visual data to enhance detection accuracy further. We then utilize a tracking module that adopts feature-based object association and the Kalman filter to track and estimate detected obstacles' states. Besides, a dynamic obstacle classification algorithm is designed to robustly identify moving objects. The dataset evaluation demonstrates a better perception performance compared to benchmark methods. The physical experiments on a quadcopter robot confirms the feasibility for real-world navigation.
no_new_dataset
0.945951
2502.20609
Mateusz Lango
J\k{e}drzej Warczy\'nski, Mateusz Lango, Ondrej Dusek
Leveraging Large Language Models for Building Interpretable Rule-Based Data-to-Text Systems
null
null
null
null
cs.CL cs.AI
http://creativecommons.org/licenses/by/4.0/
We introduce a simple approach that uses a large language model (LLM) to automatically implement a fully interpretable rule-based data-to-text system in pure Python. Experimental evaluation on the WebNLG dataset showed that such a constructed system produces text of better quality (according to the BLEU and BLEURT metrics) than the same LLM prompted to directly produce outputs, and produces fewer hallucinations than a BART language model fine-tuned on the same data. Furthermore, at runtime, the approach generates text in a fraction of the processing time required by neural approaches, using only a single CPU
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:23:55 GMT" } ]
2025-03-03T00:00:00
[ [ "Warczyński", "Jędrzej", "" ], [ "Lango", "Mateusz", "" ], [ "Dusek", "Ondrej", "" ] ]
TITLE: Leveraging Large Language Models for Building Interpretable Rule-Based Data-to-Text Systems ABSTRACT: We introduce a simple approach that uses a large language model (LLM) to automatically implement a fully interpretable rule-based data-to-text system in pure Python. Experimental evaluation on the WebNLG dataset showed that such a constructed system produces text of better quality (according to the BLEU and BLEURT metrics) than the same LLM prompted to directly produce outputs, and produces fewer hallucinations than a BART language model fine-tuned on the same data. Furthermore, at runtime, the approach generates text in a fraction of the processing time required by neural approaches, using only a single CPU
no_new_dataset
0.948728
2502.20612
Vicente Balmaseda
Vicente Balmaseda, Bokun Wang, Ching-Long Lin, Tianbao Yang
Discovering Global False Negatives On the Fly for Self-supervised Contrastive Learning
null
null
null
null
cs.LG cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In self-supervised contrastive learning, negative pairs are typically constructed using an anchor image and a sample drawn from the entire dataset, excluding the anchor. However, this approach can result in the creation of negative pairs with similar semantics, referred to as "false negatives", leading to their embeddings being falsely pushed apart. To address this issue, we introduce GloFND, an optimization-based approach that automatically learns on the fly the threshold for each anchor data to identify its false negatives during training. In contrast to previous methods for false negative discovery, our approach globally detects false negatives across the entire dataset rather than locally within the mini-batch. Moreover, its per-iteration computation cost remains independent of the dataset size. Experimental results on image and image-text data demonstrate the effectiveness of the proposed method. Our implementation is available at https://github.com/vibalcam/GloFND .
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:28:25 GMT" } ]
2025-03-03T00:00:00
[ [ "Balmaseda", "Vicente", "" ], [ "Wang", "Bokun", "" ], [ "Lin", "Ching-Long", "" ], [ "Yang", "Tianbao", "" ] ]
TITLE: Discovering Global False Negatives On the Fly for Self-supervised Contrastive Learning ABSTRACT: In self-supervised contrastive learning, negative pairs are typically constructed using an anchor image and a sample drawn from the entire dataset, excluding the anchor. However, this approach can result in the creation of negative pairs with similar semantics, referred to as "false negatives", leading to their embeddings being falsely pushed apart. To address this issue, we introduce GloFND, an optimization-based approach that automatically learns on the fly the threshold for each anchor data to identify its false negatives during training. In contrast to previous methods for false negative discovery, our approach globally detects false negatives across the entire dataset rather than locally within the mini-batch. Moreover, its per-iteration computation cost remains independent of the dataset size. Experimental results on image and image-text data demonstrate the effectiveness of the proposed method. Our implementation is available at https://github.com/vibalcam/GloFND .
no_new_dataset
0.951278
2502.20616
Juntao Tan
Juntao Tan, Liangwei Yang, Zuxin Liu, Zhiwei Liu, Rithesh Murthy, Tulika Manoj Awalgaonkar, Jianguo Zhang, Weiran Yao, Ming Zhu, Shirley Kokane, Silvio Savarese, Huan Wang, Caiming Xiong, Shelby Heinecke
PersonaBench: Evaluating AI Models on Understanding Personal Information through Accessing (Synthetic) Private User Data
null
null
null
null
cs.AI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Personalization is critical in AI assistants, particularly in the context of private AI models that work with individual users. A key scenario in this domain involves enabling AI models to access and interpret a user's private data (e.g., conversation history, user-AI interactions, app usage) to understand personal details such as biographical information, preferences, and social connections. However, due to the sensitive nature of such data, there are no publicly available datasets that allow us to assess an AI model's ability to understand users through direct access to personal information. To address this gap, we introduce a synthetic data generation pipeline that creates diverse, realistic user profiles and private documents simulating human activities. Leveraging this synthetic data, we present PersonaBench, a benchmark designed to evaluate AI models' performance in understanding personal information derived from simulated private user data. We evaluate Retrieval-Augmented Generation (RAG) pipelines using questions directly related to a user's personal information, supported by the relevant private documents provided to the models. Our results reveal that current retrieval-augmented AI models struggle to answer private questions by extracting personal information from user documents, highlighting the need for improved methodologies to enhance personalization capabilities in AI.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:43:35 GMT" } ]
2025-03-03T00:00:00
[ [ "Tan", "Juntao", "" ], [ "Yang", "Liangwei", "" ], [ "Liu", "Zuxin", "" ], [ "Liu", "Zhiwei", "" ], [ "Murthy", "Rithesh", "" ], [ "Awalgaonkar", "Tulika Manoj", "" ], [ "Zhang", "Jianguo", "" ], [ "Yao", "Weiran", "" ], [ "Zhu", "Ming", "" ], [ "Kokane", "Shirley", "" ], [ "Savarese", "Silvio", "" ], [ "Wang", "Huan", "" ], [ "Xiong", "Caiming", "" ], [ "Heinecke", "Shelby", "" ] ]
TITLE: PersonaBench: Evaluating AI Models on Understanding Personal Information through Accessing (Synthetic) Private User Data ABSTRACT: Personalization is critical in AI assistants, particularly in the context of private AI models that work with individual users. A key scenario in this domain involves enabling AI models to access and interpret a user's private data (e.g., conversation history, user-AI interactions, app usage) to understand personal details such as biographical information, preferences, and social connections. However, due to the sensitive nature of such data, there are no publicly available datasets that allow us to assess an AI model's ability to understand users through direct access to personal information. To address this gap, we introduce a synthetic data generation pipeline that creates diverse, realistic user profiles and private documents simulating human activities. Leveraging this synthetic data, we present PersonaBench, a benchmark designed to evaluate AI models' performance in understanding personal information derived from simulated private user data. We evaluate Retrieval-Augmented Generation (RAG) pipelines using questions directly related to a user's personal information, supported by the relevant private documents provided to the models. Our results reveal that current retrieval-augmented AI models struggle to answer private questions by extracting personal information from user documents, highlighting the need for improved methodologies to enhance personalization capabilities in AI.
no_new_dataset
0.576974
2502.20620
Ayana Niwa
Ayana Niwa and Masahiro Kaneko and Kentaro Inui
Rectifying Belief Space via Unlearning to Harness LLMs' Reasoning
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
Large language models (LLMs) can exhibit advanced reasoning yet still generate incorrect answers. We hypothesize that such errors frequently stem from spurious beliefs, propositions the model internally considers true but are incorrect. To address this, we propose a method to rectify the belief space by suppressing these spurious beliefs while simultaneously enhancing true ones, thereby enabling more reliable inferences. Our approach first identifies the beliefs that lead to incorrect or correct answers by prompting the model to generate textual explanations, using our Forward-Backward Beam Search (FBBS). We then apply unlearning to suppress the identified spurious beliefs and enhance the true ones, effectively rectifying the model's belief space. Empirical results on multiple QA datasets and LLMs show that our method corrects previously misanswered questions without harming overall model performance. Furthermore, our approach yields improved generalization on unseen data, suggesting that rectifying a model's belief space is a promising direction for mitigating errors and enhancing overall reliability.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 00:57:45 GMT" } ]
2025-03-03T00:00:00
[ [ "Niwa", "Ayana", "" ], [ "Kaneko", "Masahiro", "" ], [ "Inui", "Kentaro", "" ] ]
TITLE: Rectifying Belief Space via Unlearning to Harness LLMs' Reasoning ABSTRACT: Large language models (LLMs) can exhibit advanced reasoning yet still generate incorrect answers. We hypothesize that such errors frequently stem from spurious beliefs, propositions the model internally considers true but are incorrect. To address this, we propose a method to rectify the belief space by suppressing these spurious beliefs while simultaneously enhancing true ones, thereby enabling more reliable inferences. Our approach first identifies the beliefs that lead to incorrect or correct answers by prompting the model to generate textual explanations, using our Forward-Backward Beam Search (FBBS). We then apply unlearning to suppress the identified spurious beliefs and enhance the true ones, effectively rectifying the model's belief space. Empirical results on multiple QA datasets and LLMs show that our method corrects previously misanswered questions without harming overall model performance. Furthermore, our approach yields improved generalization on unseen data, suggesting that rectifying a model's belief space is a promising direction for mitigating errors and enhancing overall reliability.
no_new_dataset
0.952353
2502.20622
Chi Ruan
Chi Ruan
RTGen: Real-Time Generative Detection Transformer
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
While open-vocabulary object detectors require predefined categories during inference, generative object detectors overcome this limitation by endowing the model with text generation capabilities. However, existing generative object detection methods directly append an autoregressive language model to an object detector to generate texts for each detected object. This straightforward design leads to structural redundancy and increased processing time. In this paper, we propose a Real-Time GENerative Detection Transformer (RTGen), a real-time generative object detector with a succinct encoder-decoder architecture. Specifically, we introduce a novel Region-Language Decoder (RL-Decoder), which innovatively integrates a non-autoregressive language model into the detection decoder, enabling concurrent processing of object and text information. With these efficient designs, RTGen achieves a remarkable inference speed of 60.41 FPS. Moreover, RTGen obtains 18.6 mAP on the LVIS dataset, outperforming the previous SOTA method by 3.5 mAP.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 01:01:56 GMT" } ]
2025-03-03T00:00:00
[ [ "Ruan", "Chi", "" ] ]
TITLE: RTGen: Real-Time Generative Detection Transformer ABSTRACT: While open-vocabulary object detectors require predefined categories during inference, generative object detectors overcome this limitation by endowing the model with text generation capabilities. However, existing generative object detection methods directly append an autoregressive language model to an object detector to generate texts for each detected object. This straightforward design leads to structural redundancy and increased processing time. In this paper, we propose a Real-Time GENerative Detection Transformer (RTGen), a real-time generative object detector with a succinct encoder-decoder architecture. Specifically, we introduce a novel Region-Language Decoder (RL-Decoder), which innovatively integrates a non-autoregressive language model into the detection decoder, enabling concurrent processing of object and text information. With these efficient designs, RTGen achieves a remarkable inference speed of 60.41 FPS. Moreover, RTGen obtains 18.6 mAP on the LVIS dataset, outperforming the previous SOTA method by 3.5 mAP.
no_new_dataset
0.946892
2502.20623
Yuepeng Hu
Yuepeng Hu, Zhengyuan Jiang, Neil Zhenqiang Gong
SafeText: Safe Text-to-image Models via Aligning the Text Encoder
null
null
null
null
cs.CR cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Text-to-image models can generate harmful images when presented with unsafe prompts, posing significant safety and societal risks. Alignment methods aim to modify these models to ensure they generate only non-harmful images, even when exposed to unsafe prompts. A typical text-to-image model comprises two main components: 1) a text encoder and 2) a diffusion module. Existing alignment methods mainly focus on modifying the diffusion module to prevent harmful image generation. However, this often significantly impacts the model's behavior for safe prompts, causing substantial quality degradation of generated images. In this work, we propose SafeText, a novel alignment method that fine-tunes the text encoder rather than the diffusion module. By adjusting the text encoder, SafeText significantly alters the embedding vectors for unsafe prompts, while minimally affecting those for safe prompts. As a result, the diffusion module generates non-harmful images for unsafe prompts while preserving the quality of images for safe prompts. We evaluate SafeText on multiple datasets of safe and unsafe prompts, including those generated through jailbreak attacks. Our results show that SafeText effectively prevents harmful image generation with minor impact on the images for safe prompts, and SafeText outperforms six existing alignment methods. We will publish our code and data after paper acceptance.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 01:02:57 GMT" } ]
2025-03-03T00:00:00
[ [ "Hu", "Yuepeng", "" ], [ "Jiang", "Zhengyuan", "" ], [ "Gong", "Neil Zhenqiang", "" ] ]
TITLE: SafeText: Safe Text-to-image Models via Aligning the Text Encoder ABSTRACT: Text-to-image models can generate harmful images when presented with unsafe prompts, posing significant safety and societal risks. Alignment methods aim to modify these models to ensure they generate only non-harmful images, even when exposed to unsafe prompts. A typical text-to-image model comprises two main components: 1) a text encoder and 2) a diffusion module. Existing alignment methods mainly focus on modifying the diffusion module to prevent harmful image generation. However, this often significantly impacts the model's behavior for safe prompts, causing substantial quality degradation of generated images. In this work, we propose SafeText, a novel alignment method that fine-tunes the text encoder rather than the diffusion module. By adjusting the text encoder, SafeText significantly alters the embedding vectors for unsafe prompts, while minimally affecting those for safe prompts. As a result, the diffusion module generates non-harmful images for unsafe prompts while preserving the quality of images for safe prompts. We evaluate SafeText on multiple datasets of safe and unsafe prompts, including those generated through jailbreak attacks. Our results show that SafeText effectively prevents harmful image generation with minor impact on the images for safe prompts, and SafeText outperforms six existing alignment methods. We will publish our code and data after paper acceptance.
no_new_dataset
0.949856
2502.20639
Leming Shen
Leming Shen, Qiang Yang, Kaiyan Cui, Yuanqing Zheng, Xiao-Yong Wei, Jianwei Liu, Jinsong Han
FedConv: A Learning-on-Model Paradigm for Heterogeneous Federated Clients
null
null
null
null
cs.LG cs.AI
http://creativecommons.org/publicdomain/zero/1.0/
Federated Learning (FL) facilitates collaborative training of a shared global model without exposing clients' private data. In practical FL systems, clients (e.g., edge servers, smartphones, and wearables) typically have disparate system resources. Conventional FL, however, adopts a one-size-fits-all solution, where a homogeneous large global model is transmitted to and trained on each client, resulting in an overwhelming workload for less capable clients and starvation for other clients. To address this issue, we propose FedConv, a client-friendly FL framework, which minimizes the computation and memory burden on resource-constrained clients by providing heterogeneous customized sub-models. FedConv features a novel learning-on-model paradigm that learns the parameters of the heterogeneous sub-models via convolutional compression. Unlike traditional compression methods, the compressed models in FedConv can be directly trained on clients without decompression. To aggregate the heterogeneous sub-models, we propose transposed convolutional dilation to convert them back to large models with a unified size while retaining personalized information from clients. The compression and dilation processes, transparent to clients, are optimized on the server leveraging a small public dataset. Extensive experiments on six datasets demonstrate that FedConv outperforms state-of-the-art FL systems in terms of model accuracy (by more than 35% on average), computation and communication overhead (with 33% and 25% reduction, respectively).
[ { "version": "v1", "created": "Fri, 28 Feb 2025 01:39:53 GMT" } ]
2025-03-03T00:00:00
[ [ "Shen", "Leming", "" ], [ "Yang", "Qiang", "" ], [ "Cui", "Kaiyan", "" ], [ "Zheng", "Yuanqing", "" ], [ "Wei", "Xiao-Yong", "" ], [ "Liu", "Jianwei", "" ], [ "Han", "Jinsong", "" ] ]
TITLE: FedConv: A Learning-on-Model Paradigm for Heterogeneous Federated Clients ABSTRACT: Federated Learning (FL) facilitates collaborative training of a shared global model without exposing clients' private data. In practical FL systems, clients (e.g., edge servers, smartphones, and wearables) typically have disparate system resources. Conventional FL, however, adopts a one-size-fits-all solution, where a homogeneous large global model is transmitted to and trained on each client, resulting in an overwhelming workload for less capable clients and starvation for other clients. To address this issue, we propose FedConv, a client-friendly FL framework, which minimizes the computation and memory burden on resource-constrained clients by providing heterogeneous customized sub-models. FedConv features a novel learning-on-model paradigm that learns the parameters of the heterogeneous sub-models via convolutional compression. Unlike traditional compression methods, the compressed models in FedConv can be directly trained on clients without decompression. To aggregate the heterogeneous sub-models, we propose transposed convolutional dilation to convert them back to large models with a unified size while retaining personalized information from clients. The compression and dilation processes, transparent to clients, are optimized on the server leveraging a small public dataset. Extensive experiments on six datasets demonstrate that FedConv outperforms state-of-the-art FL systems in terms of model accuracy (by more than 35% on average), computation and communication overhead (with 33% and 25% reduction, respectively).
no_new_dataset
0.951097
2502.20643
Pengyu Zhang
Pengyu Zhang, Xieyuanli Chen, Yuwei Chen, Beizhen Bi, Zhuo Xu, Tian Jin, Xiaotao Huang, Liang Shen
EDENet: Echo Direction Encoding Network for Place Recognition Based on Ground Penetrating Radar
null
null
null
null
cs.CV cs.RO
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Ground penetrating radar (GPR) based localization has gained significant recognition in robotics due to its ability to detect stable subsurface features, offering advantages in environments where traditional sensors like cameras and LiDAR may struggle. However, existing methods are primarily focused on small-scale place recognition (PR), leaving the challenges of PR in large-scale maps unaddressed. These challenges include the inherent sparsity of underground features and the variability in underground dielectric constants, which complicate robust localization. In this work, we investigate the geometric relationship between GPR echo sequences and underground scenes, leveraging the robustness of directional features to inform our network design. We introduce learnable Gabor filters for the precise extraction of directional responses, coupled with a direction-aware attention mechanism for effective geometric encoding. To further enhance performance, we incorporate a shift-invariant unit and a multi-scale aggregation strategy to better accommodate variations in di-electric constants. Experiments conducted on public datasets demonstrate that our proposed EDENet not only surpasses existing solutions in terms of PR performance but also offers advantages in model size and computational efficiency.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 01:48:12 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhang", "Pengyu", "" ], [ "Chen", "Xieyuanli", "" ], [ "Chen", "Yuwei", "" ], [ "Bi", "Beizhen", "" ], [ "Xu", "Zhuo", "" ], [ "Jin", "Tian", "" ], [ "Huang", "Xiaotao", "" ], [ "Shen", "Liang", "" ] ]
TITLE: EDENet: Echo Direction Encoding Network for Place Recognition Based on Ground Penetrating Radar ABSTRACT: Ground penetrating radar (GPR) based localization has gained significant recognition in robotics due to its ability to detect stable subsurface features, offering advantages in environments where traditional sensors like cameras and LiDAR may struggle. However, existing methods are primarily focused on small-scale place recognition (PR), leaving the challenges of PR in large-scale maps unaddressed. These challenges include the inherent sparsity of underground features and the variability in underground dielectric constants, which complicate robust localization. In this work, we investigate the geometric relationship between GPR echo sequences and underground scenes, leveraging the robustness of directional features to inform our network design. We introduce learnable Gabor filters for the precise extraction of directional responses, coupled with a direction-aware attention mechanism for effective geometric encoding. To further enhance performance, we incorporate a shift-invariant unit and a multi-scale aggregation strategy to better accommodate variations in di-electric constants. Experiments conducted on public datasets demonstrate that our proposed EDENet not only surpasses existing solutions in terms of PR performance but also offers advantages in model size and computational efficiency.
no_new_dataset
0.950778
2502.20647
Haleh Shahzad Dr
Colleen Gilhuly, Haleh Shahzad
Consistency Evaluation of News Article Summaries Generated by Large (and Small) Language Models
21 pages, 6 figures, 4 tables
null
null
null
cs.CL cs.AI cs.LG
http://creativecommons.org/licenses/by/4.0/
Text summarizing is a critical Natural Language Processing (NLP) task with applications ranging from information retrieval to content generation. Large Language Models (LLMs) have shown remarkable promise in generating fluent abstractive summaries but they can produce hallucinated details not grounded in the source text. Regardless of the method of generating a summary, high quality automated evaluations remain an open area of investigation. This paper embarks on an exploration of text summarization with a diverse set of techniques, including TextRank, BART, Mistral-7B-Instruct, and OpenAI GPT-3.5-Turbo. The generated summaries are evaluated using traditional metrics such as the Recall-Oriented Understudy for Gisting Evaluation (ROUGE) Score and Bidirectional Encoder Representations from Transformers (BERT) Score, as well as LLM-powered evaluation methods that directly assess a generated summary's consistency with the source text. We introduce a meta evaluation score which directly assesses the performance of the LLM evaluation system (prompt + model). We find that that all summarization models produce consistent summaries when tested on the XL-Sum dataset, exceeding the consistency of the reference summaries.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 01:58:17 GMT" } ]
2025-03-03T00:00:00
[ [ "Gilhuly", "Colleen", "" ], [ "Shahzad", "Haleh", "" ] ]
TITLE: Consistency Evaluation of News Article Summaries Generated by Large (and Small) Language Models ABSTRACT: Text summarizing is a critical Natural Language Processing (NLP) task with applications ranging from information retrieval to content generation. Large Language Models (LLMs) have shown remarkable promise in generating fluent abstractive summaries but they can produce hallucinated details not grounded in the source text. Regardless of the method of generating a summary, high quality automated evaluations remain an open area of investigation. This paper embarks on an exploration of text summarization with a diverse set of techniques, including TextRank, BART, Mistral-7B-Instruct, and OpenAI GPT-3.5-Turbo. The generated summaries are evaluated using traditional metrics such as the Recall-Oriented Understudy for Gisting Evaluation (ROUGE) Score and Bidirectional Encoder Representations from Transformers (BERT) Score, as well as LLM-powered evaluation methods that directly assess a generated summary's consistency with the source text. We introduce a meta evaluation score which directly assesses the performance of the LLM evaluation system (prompt + model). We find that that all summarization models produce consistent summaries when tested on the XL-Sum dataset, exceeding the consistency of the reference summaries.
no_new_dataset
0.935582
2502.20651
Sakshi Singh
Rishi Mukherjee, Sakshi Singh, Jack McWilliams, Junaed Sattar
The Common Objects Underwater (COU) Dataset for Robust Underwater Object Detection
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
We introduce COU: Common Objects Underwater, an instance-segmented image dataset of commonly found man-made objects in multiple aquatic and marine environments. COU contains approximately 10K segmented images, annotated from images collected during a number of underwater robot field trials in diverse locations. COU has been created to address the lack of datasets with robust class coverage curated for underwater instance segmentation, which is particularly useful for training light-weight, real-time capable detectors for Autonomous Underwater Vehicles (AUVs). In addition, COU addresses the lack of diversity in object classes since the commonly available underwater image datasets focus only on marine life. Currently, COU contains images from both closed-water (pool) and open-water (lakes and oceans) environments, of 24 different classes of objects including marine debris, dive tools, and AUVs. To assess the efficacy of COU in training underwater object detectors, we use three state-of-the-art models to evaluate its performance and accuracy, using a combination of standard accuracy and efficiency metrics. The improved performance of COU-trained detectors over those solely trained on terrestrial data demonstrates the clear advantage of training with annotated underwater images. We make COU available for broad use under open-source licenses.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 02:12:24 GMT" } ]
2025-03-03T00:00:00
[ [ "Mukherjee", "Rishi", "" ], [ "Singh", "Sakshi", "" ], [ "McWilliams", "Jack", "" ], [ "Sattar", "Junaed", "" ] ]
TITLE: The Common Objects Underwater (COU) Dataset for Robust Underwater Object Detection ABSTRACT: We introduce COU: Common Objects Underwater, an instance-segmented image dataset of commonly found man-made objects in multiple aquatic and marine environments. COU contains approximately 10K segmented images, annotated from images collected during a number of underwater robot field trials in diverse locations. COU has been created to address the lack of datasets with robust class coverage curated for underwater instance segmentation, which is particularly useful for training light-weight, real-time capable detectors for Autonomous Underwater Vehicles (AUVs). In addition, COU addresses the lack of diversity in object classes since the commonly available underwater image datasets focus only on marine life. Currently, COU contains images from both closed-water (pool) and open-water (lakes and oceans) environments, of 24 different classes of objects including marine debris, dive tools, and AUVs. To assess the efficacy of COU in training underwater object detectors, we use three state-of-the-art models to evaluate its performance and accuracy, using a combination of standard accuracy and efficiency metrics. The improved performance of COU-trained detectors over those solely trained on terrestrial data demonstrates the clear advantage of training with annotated underwater images. We make COU available for broad use under open-source licenses.
new_dataset
0.97151
2502.20653
Shaobo Wang
Shaobo Wang and Yicun Yang and Zhiyuan Liu and Chenghao Sun and Xuming Hu and Conghui He and Linfeng Zhang
Dataset Distillation with Neural Characteristic Function: A Minmax Perspective
Accepted by CVPR 2025, 11 pages, 7 figures
Conference on Computer Vision and Pattern Recognition, 2025
null
null
cs.CV cs.AI cs.LG
http://creativecommons.org/licenses/by/4.0/
Dataset distillation has emerged as a powerful approach for reducing data requirements in deep learning. Among various methods, distribution matching-based approaches stand out for their balance of computational efficiency and strong performance. However, existing distance metrics used in distribution matching often fail to accurately capture distributional differences, leading to unreliable measures of discrepancy. In this paper, we reformulate dataset distillation as a minmax optimization problem and introduce Neural Characteristic Function Discrepancy (NCFD), a comprehensive and theoretically grounded metric for measuring distributional differences. NCFD leverages the Characteristic Function (CF) to encapsulate full distributional information, employing a neural network to optimize the sampling strategy for the CF's frequency arguments, thereby maximizing the discrepancy to enhance distance estimation. Simultaneously, we minimize the difference between real and synthetic data under this optimized NCFD measure. Our approach, termed Neural Characteristic Function Matching (\mymethod{}), inherently aligns the phase and amplitude of neural features in the complex plane for both real and synthetic data, achieving a balance between realism and diversity in synthetic samples. Experiments demonstrate that our method achieves significant performance gains over state-of-the-art methods on both low- and high-resolution datasets. Notably, we achieve a 20.5\% accuracy boost on ImageSquawk. Our method also reduces GPU memory usage by over 300$\times$ and achieves 20$\times$ faster processing speeds compared to state-of-the-art methods. To the best of our knowledge, this is the first work to achieve lossless compression of CIFAR-100 on a single NVIDIA 2080 Ti GPU using only 2.3 GB of memory.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 02:14:55 GMT" } ]
2025-03-03T00:00:00
[ [ "Wang", "Shaobo", "" ], [ "Yang", "Yicun", "" ], [ "Liu", "Zhiyuan", "" ], [ "Sun", "Chenghao", "" ], [ "Hu", "Xuming", "" ], [ "He", "Conghui", "" ], [ "Zhang", "Linfeng", "" ] ]
TITLE: Dataset Distillation with Neural Characteristic Function: A Minmax Perspective ABSTRACT: Dataset distillation has emerged as a powerful approach for reducing data requirements in deep learning. Among various methods, distribution matching-based approaches stand out for their balance of computational efficiency and strong performance. However, existing distance metrics used in distribution matching often fail to accurately capture distributional differences, leading to unreliable measures of discrepancy. In this paper, we reformulate dataset distillation as a minmax optimization problem and introduce Neural Characteristic Function Discrepancy (NCFD), a comprehensive and theoretically grounded metric for measuring distributional differences. NCFD leverages the Characteristic Function (CF) to encapsulate full distributional information, employing a neural network to optimize the sampling strategy for the CF's frequency arguments, thereby maximizing the discrepancy to enhance distance estimation. Simultaneously, we minimize the difference between real and synthetic data under this optimized NCFD measure. Our approach, termed Neural Characteristic Function Matching (\mymethod{}), inherently aligns the phase and amplitude of neural features in the complex plane for both real and synthetic data, achieving a balance between realism and diversity in synthetic samples. Experiments demonstrate that our method achieves significant performance gains over state-of-the-art methods on both low- and high-resolution datasets. Notably, we achieve a 20.5\% accuracy boost on ImageSquawk. Our method also reduces GPU memory usage by over 300$\times$ and achieves 20$\times$ faster processing speeds compared to state-of-the-art methods. To the best of our knowledge, this is the first work to achieve lossless compression of CIFAR-100 on a single NVIDIA 2080 Ti GPU using only 2.3 GB of memory.
no_new_dataset
0.949295
2502.20661
Chaeyun Jang
Hyungi Lee, Chaeyun Jang, Dongbok Lee, Juho Lee
Dimension Agnostic Neural Processes
10 pages, 5 figures, Accepted to ICLR 2025 (International Conference on Learning Representations)
null
null
null
cs.LG
http://creativecommons.org/licenses/by/4.0/
Meta-learning aims to train models that can generalize to new tasks with limited labeled data by extracting shared features across diverse task datasets. Additionally, it accounts for prediction uncertainty during both training and evaluation, a concept known as uncertainty-aware meta-learning. Neural Process(NP) is a well-known uncertainty-aware meta-learning method that constructs implicit stochastic processes using parametric neural networks, enabling rapid adaptation to new tasks. However, existing NP methods face challenges in accommodating diverse input dimensions and learned features, limiting their broad applicability across regression tasks. To address these limitations and advance the utility of NP models as general regressors, we introduce Dimension Agnostic Neural Processes(DANP). DANP incorporates Dimension Aggregator Block(DAB) to transform input features into a fixed-dimensional space, enhancing the model's ability to handle diverse datasets. Furthermore, leveraging the Transformer architecture and latent encoding layers, DANP learns a wider range of features that are generalizable across various tasks. Through comprehensive experimentation on various synthetic and practical regression tasks, we empirically show that DANP outperforms previous NP variations, showcasing its effectiveness in overcoming the limitations of traditional NP models and its potential for broader applicability in diverse regression scenarios.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 02:40:59 GMT" } ]
2025-03-03T00:00:00
[ [ "Lee", "Hyungi", "" ], [ "Jang", "Chaeyun", "" ], [ "Lee", "Dongbok", "" ], [ "Lee", "Juho", "" ] ]
TITLE: Dimension Agnostic Neural Processes ABSTRACT: Meta-learning aims to train models that can generalize to new tasks with limited labeled data by extracting shared features across diverse task datasets. Additionally, it accounts for prediction uncertainty during both training and evaluation, a concept known as uncertainty-aware meta-learning. Neural Process(NP) is a well-known uncertainty-aware meta-learning method that constructs implicit stochastic processes using parametric neural networks, enabling rapid adaptation to new tasks. However, existing NP methods face challenges in accommodating diverse input dimensions and learned features, limiting their broad applicability across regression tasks. To address these limitations and advance the utility of NP models as general regressors, we introduce Dimension Agnostic Neural Processes(DANP). DANP incorporates Dimension Aggregator Block(DAB) to transform input features into a fixed-dimensional space, enhancing the model's ability to handle diverse datasets. Furthermore, leveraging the Transformer architecture and latent encoding layers, DANP learns a wider range of features that are generalizable across various tasks. Through comprehensive experimentation on various synthetic and practical regression tasks, we empirically show that DANP outperforms previous NP variations, showcasing its effectiveness in overcoming the limitations of traditional NP models and its potential for broader applicability in diverse regression scenarios.
no_new_dataset
0.944434
2502.20667
Ojonugwa Ejiga Peter
Ojonugwa Oluwafemi Ejiga Peter, Md Mahmudur Rahman, and Fahmi Khalifa
Advancing AI-Powered Medical Image Synthesis: Insights from MedVQA-GI Challenge Using CLIP, Fine-Tuned Stable Diffusion, and Dream-Booth + LoRA
null
null
null
null
cs.CV cs.AI cs.LG
http://creativecommons.org/licenses/by/4.0/
The MEDVQA-GI challenge addresses the integration of AI-driven text-to-image generative models in medical diagnostics, aiming to enhance diagnostic capabilities through synthetic image generation. Existing methods primarily focus on static image analysis and lack the dynamic generation of medical imagery from textual descriptions. This study intends to partially close this gap by introducing a novel approach based on fine-tuned generative models to generate dynamic, scalable, and precise images from textual descriptions. Particularly, our system integrates fine-tuned Stable Diffusion and DreamBooth models, as well as Low-Rank Adaptation (LORA), to generate high-fidelity medical images. The problem is around two sub-tasks namely: image synthesis (IS) and optimal prompt production (OPG). The former creates medical images via verbal prompts, whereas the latter provides prompts that produce high-quality images in specified categories. The study emphasizes the limitations of traditional medical image generation methods, such as hand sketching, constrained datasets, static procedures, and generic models. Our evaluation measures showed that Stable Diffusion surpasses CLIP and DreamBooth + LORA in terms of producing high-quality, diversified images. Specifically, Stable Diffusion had the lowest Fr\'echet Inception Distance (FID) scores (0.099 for single center, 0.064 for multi-center, and 0.067 for combined), indicating higher image quality. Furthermore, it had the highest average Inception Score (2.327 across all datasets), indicating exceptional diversity and quality. This advances the field of AI-powered medical diagnosis. Future research will concentrate on model refining, dataset augmentation, and ethical considerations for efficiently implementing these advances into clinical practice
[ { "version": "v1", "created": "Fri, 28 Feb 2025 02:49:45 GMT" } ]
2025-03-03T00:00:00
[ [ "Peter", "Ojonugwa Oluwafemi Ejiga", "" ], [ "Rahman", "Md Mahmudur", "" ], [ "Khalifa", "Fahmi", "" ] ]
TITLE: Advancing AI-Powered Medical Image Synthesis: Insights from MedVQA-GI Challenge Using CLIP, Fine-Tuned Stable Diffusion, and Dream-Booth + LoRA ABSTRACT: The MEDVQA-GI challenge addresses the integration of AI-driven text-to-image generative models in medical diagnostics, aiming to enhance diagnostic capabilities through synthetic image generation. Existing methods primarily focus on static image analysis and lack the dynamic generation of medical imagery from textual descriptions. This study intends to partially close this gap by introducing a novel approach based on fine-tuned generative models to generate dynamic, scalable, and precise images from textual descriptions. Particularly, our system integrates fine-tuned Stable Diffusion and DreamBooth models, as well as Low-Rank Adaptation (LORA), to generate high-fidelity medical images. The problem is around two sub-tasks namely: image synthesis (IS) and optimal prompt production (OPG). The former creates medical images via verbal prompts, whereas the latter provides prompts that produce high-quality images in specified categories. The study emphasizes the limitations of traditional medical image generation methods, such as hand sketching, constrained datasets, static procedures, and generic models. Our evaluation measures showed that Stable Diffusion surpasses CLIP and DreamBooth + LORA in terms of producing high-quality, diversified images. Specifically, Stable Diffusion had the lowest Fr\'echet Inception Distance (FID) scores (0.099 for single center, 0.064 for multi-center, and 0.067 for combined), indicating higher image quality. Furthermore, it had the highest average Inception Score (2.327 across all datasets), indicating exceptional diversity and quality. This advances the field of AI-powered medical diagnosis. Future research will concentrate on model refining, dataset augmentation, and ethical considerations for efficiently implementing these advances into clinical practice
no_new_dataset
0.955068
2502.20668
Xiang Xiang
Xiang Xiang, Zhuo Xu, Yao Deng, Qinhao Zhou, Yifan Liang, Ke Chen, Qingfang Zheng, Yaowei Wang, Xilin Chen, Wen Gao
OpenEarthSensing: Large-Scale Fine-Grained Benchmark for Open-World Remote Sensing
null
null
null
null
cs.CV cs.AI cs.LG eess.IV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In open-world remote sensing, deployed models must continuously adapt to a steady influx of new data, which often exhibits various shifts compared to what the model encountered during the training phase. To effectively handle the new data, models are required to detect semantic shifts, adapt to covariate shifts, and continuously update themselves. These challenges give rise to a variety of open-world tasks. However, existing open-world remote sensing studies typically train and test within a single dataset to simulate open-world conditions. Currently, there is a lack of large-scale benchmarks capable of evaluating multiple open-world tasks. In this paper, we introduce OpenEarthSensing, a large-scale fine-grained benchmark for open-world remote sensing. OpenEarthSensing includes 189 scene and objects categories, covering the vast majority of potential semantic shifts that may occur in the real world. Additionally, OpenEarthSensing encompasses five data domains with significant covariate shifts, including two RGB satellite domians, one RGB aerial domian, one MS RGB domian, and one infrared domian. The various domains provide a more comprehensive testbed for evaluating the generalization performance of open-world models. We conduct the baseline evaluation of current mainstream open-world tasks and methods on OpenEarthSensing, demonstrating that it serves as a challenging benchmark for open-world remote sensing.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 02:49:52 GMT" } ]
2025-03-03T00:00:00
[ [ "Xiang", "Xiang", "" ], [ "Xu", "Zhuo", "" ], [ "Deng", "Yao", "" ], [ "Zhou", "Qinhao", "" ], [ "Liang", "Yifan", "" ], [ "Chen", "Ke", "" ], [ "Zheng", "Qingfang", "" ], [ "Wang", "Yaowei", "" ], [ "Chen", "Xilin", "" ], [ "Gao", "Wen", "" ] ]
TITLE: OpenEarthSensing: Large-Scale Fine-Grained Benchmark for Open-World Remote Sensing ABSTRACT: In open-world remote sensing, deployed models must continuously adapt to a steady influx of new data, which often exhibits various shifts compared to what the model encountered during the training phase. To effectively handle the new data, models are required to detect semantic shifts, adapt to covariate shifts, and continuously update themselves. These challenges give rise to a variety of open-world tasks. However, existing open-world remote sensing studies typically train and test within a single dataset to simulate open-world conditions. Currently, there is a lack of large-scale benchmarks capable of evaluating multiple open-world tasks. In this paper, we introduce OpenEarthSensing, a large-scale fine-grained benchmark for open-world remote sensing. OpenEarthSensing includes 189 scene and objects categories, covering the vast majority of potential semantic shifts that may occur in the real world. Additionally, OpenEarthSensing encompasses five data domains with significant covariate shifts, including two RGB satellite domians, one RGB aerial domian, one MS RGB domian, and one infrared domian. The various domains provide a more comprehensive testbed for evaluating the generalization performance of open-world models. We conduct the baseline evaluation of current mainstream open-world tasks and methods on OpenEarthSensing, demonstrating that it serves as a challenging benchmark for open-world remote sensing.
no_new_dataset
0.909667
2502.20669
John Han
John J. Han, Jie Ying Wu
EndoPBR: Material and Lighting Estimation for Photorealistic Surgical Simulations via Physically-based Rendering
10 pages, 3 figures
null
null
null
cs.CV
http://creativecommons.org/licenses/by-nc-sa/4.0/
The lack of labeled datasets in 3D vision for surgical scenes inhibits the development of robust 3D reconstruction algorithms in the medical domain. Despite the popularity of Neural Radiance Fields and 3D Gaussian Splatting in the general computer vision community, these systems have yet to find consistent success in surgical scenes due to challenges such as non-stationary lighting and non-Lambertian surfaces. As a result, the need for labeled surgical datasets continues to grow. In this work, we introduce a differentiable rendering framework for material and lighting estimation from endoscopic images and known geometry. Compared to previous approaches that model lighting and material jointly as radiance, we explicitly disentangle these scene properties for robust and photorealistic novel view synthesis. To disambiguate the training process, we formulate domain-specific properties inherent in surgical scenes. Specifically, we model the scene lighting as a simple spotlight and material properties as a bidirectional reflectance distribution function, parameterized by a neural network. By grounding color predictions in the rendering equation, we can generate photorealistic images at arbitrary camera poses. We evaluate our method with various sequences from the Colonoscopy 3D Video Dataset and show that our method produces competitive novel view synthesis results compared with other approaches. Furthermore, we demonstrate that synthetic data can be used to develop 3D vision algorithms by finetuning a depth estimation model with our rendered outputs. Overall, we see that the depth estimation performance is on par with fine-tuning with the original real images.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 02:50:59 GMT" } ]
2025-03-03T00:00:00
[ [ "Han", "John J.", "" ], [ "Wu", "Jie Ying", "" ] ]
TITLE: EndoPBR: Material and Lighting Estimation for Photorealistic Surgical Simulations via Physically-based Rendering ABSTRACT: The lack of labeled datasets in 3D vision for surgical scenes inhibits the development of robust 3D reconstruction algorithms in the medical domain. Despite the popularity of Neural Radiance Fields and 3D Gaussian Splatting in the general computer vision community, these systems have yet to find consistent success in surgical scenes due to challenges such as non-stationary lighting and non-Lambertian surfaces. As a result, the need for labeled surgical datasets continues to grow. In this work, we introduce a differentiable rendering framework for material and lighting estimation from endoscopic images and known geometry. Compared to previous approaches that model lighting and material jointly as radiance, we explicitly disentangle these scene properties for robust and photorealistic novel view synthesis. To disambiguate the training process, we formulate domain-specific properties inherent in surgical scenes. Specifically, we model the scene lighting as a simple spotlight and material properties as a bidirectional reflectance distribution function, parameterized by a neural network. By grounding color predictions in the rendering equation, we can generate photorealistic images at arbitrary camera poses. We evaluate our method with various sequences from the Colonoscopy 3D Video Dataset and show that our method produces competitive novel view synthesis results compared with other approaches. Furthermore, we demonstrate that synthetic data can be used to develop 3D vision algorithms by finetuning a depth estimation model with our rendered outputs. Overall, we see that the depth estimation performance is on par with fine-tuning with the original real images.
no_new_dataset
0.951953
2502.20676
Shanshan Wan
Shanshan Wan and Yingmei Wei and Lai Kang and Tianrui Shen and Haixuan Wang and Yee-Hong Yang
SciceVPR: Stable Cross-Image Correlation Enhanced Model for Visual Place Recognition
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Visual Place Recognition (VPR) is a major challenge for robotics and autonomous systems, with the goal of predicting the location of an image based solely on its visual features. State-of-the-art (SOTA) models extract global descriptors using the powerful foundation model DINOv2 as backbone. These models either explore the cross-image correlation or propose a time-consuming two-stage re-ranking strategy to achieve better performance. However, existing works only utilize the final output of DINOv2, and the current cross-image correlation causes unstable retrieval results. To produce both discriminative and constant global descriptors, this paper proposes stable cross-image correlation enhanced model for VPR called SciceVPR. This model explores the full potential of DINOv2 in providing useful feature representations that implicitly encode valuable contextual knowledge. Specifically, SciceVPR first uses a multi-layer feature fusion module to capture increasingly detailed task-relevant channel and spatial information from the multi-layer output of DINOv2. Secondly, SciceVPR considers the invariant correlation between images within a batch as valuable knowledge to be distilled into the proposed self-enhanced encoder. In this way, SciceVPR can acquire fairly robust global features regardless of domain shifts (e.g., changes in illumination, weather and viewpoint between pictures taken in the same place). Experimental results demonstrate that the base variant, SciceVPR-B, outperforms SOTA one-stage methods with single input on multiple datasets with varying domain conditions. The large variant, SciceVPR-L, performs on par with SOTA two-stage models, scoring over 3% higher in Recall@1 compared to existing models on the challenging Tokyo24/7 dataset. Our code will be released at https://github.com/shuimushan/SciceVPR.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:05:30 GMT" } ]
2025-03-03T00:00:00
[ [ "Wan", "Shanshan", "" ], [ "Wei", "Yingmei", "" ], [ "Kang", "Lai", "" ], [ "Shen", "Tianrui", "" ], [ "Wang", "Haixuan", "" ], [ "Yang", "Yee-Hong", "" ] ]
TITLE: SciceVPR: Stable Cross-Image Correlation Enhanced Model for Visual Place Recognition ABSTRACT: Visual Place Recognition (VPR) is a major challenge for robotics and autonomous systems, with the goal of predicting the location of an image based solely on its visual features. State-of-the-art (SOTA) models extract global descriptors using the powerful foundation model DINOv2 as backbone. These models either explore the cross-image correlation or propose a time-consuming two-stage re-ranking strategy to achieve better performance. However, existing works only utilize the final output of DINOv2, and the current cross-image correlation causes unstable retrieval results. To produce both discriminative and constant global descriptors, this paper proposes stable cross-image correlation enhanced model for VPR called SciceVPR. This model explores the full potential of DINOv2 in providing useful feature representations that implicitly encode valuable contextual knowledge. Specifically, SciceVPR first uses a multi-layer feature fusion module to capture increasingly detailed task-relevant channel and spatial information from the multi-layer output of DINOv2. Secondly, SciceVPR considers the invariant correlation between images within a batch as valuable knowledge to be distilled into the proposed self-enhanced encoder. In this way, SciceVPR can acquire fairly robust global features regardless of domain shifts (e.g., changes in illumination, weather and viewpoint between pictures taken in the same place). Experimental results demonstrate that the base variant, SciceVPR-B, outperforms SOTA one-stage methods with single input on multiple datasets with varying domain conditions. The large variant, SciceVPR-L, performs on par with SOTA two-stage models, scoring over 3% higher in Recall@1 compared to existing models on the challenging Tokyo24/7 dataset. Our code will be released at https://github.com/shuimushan/SciceVPR.
no_new_dataset
0.945045
2502.20677
Youbing Hu
Youbing Hu, Yun Cheng, Zimu Zhou, Anqi Lu, Zhiqiang Cao, Zhijun Li
FoCTTA: Low-Memory Continual Test-Time Adaptation with Focus
null
null
null
null
cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Continual adaptation to domain shifts at test time (CTTA) is crucial for enhancing the intelligence of deep learning enabled IoT applications. However, prevailing TTA methods, which typically update all batch normalization (BN) layers, exhibit two memory inefficiencies. First, the reliance on BN layers for adaptation necessitates large batch sizes, leading to high memory usage. Second, updating all BN layers requires storing the activations of all BN layers for backpropagation, exacerbating the memory demand. Both factors lead to substantial memory costs, making existing solutions impractical for IoT devices. In this paper, we present FoCTTA, a low-memory CTTA strategy. The key is to automatically identify and adapt a few drift-sensitive representation layers, rather than blindly update all BN layers. The shift from BN to representation layers eliminates the need for large batch sizes. Also, by updating adaptation-critical layers only, FoCTTA avoids storing excessive activations. This focused adaptation approach ensures that FoCTTA is not only memory-efficient but also maintains effective adaptation. Evaluations show that FoCTTA improves the adaptation accuracy over the state-of-the-arts by 4.5%, 4.9%, and 14.8% on CIFAR10-C, CIFAR100-C, and ImageNet-C under the same memory constraints. Across various batch sizes, FoCTTA reduces the memory usage by 3-fold on average, while improving the accuracy by 8.1%, 3.6%, and 0.2%, respectively, on the three datasets.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:06:15 GMT" } ]
2025-03-03T00:00:00
[ [ "Hu", "Youbing", "" ], [ "Cheng", "Yun", "" ], [ "Zhou", "Zimu", "" ], [ "Lu", "Anqi", "" ], [ "Cao", "Zhiqiang", "" ], [ "Li", "Zhijun", "" ] ]
TITLE: FoCTTA: Low-Memory Continual Test-Time Adaptation with Focus ABSTRACT: Continual adaptation to domain shifts at test time (CTTA) is crucial for enhancing the intelligence of deep learning enabled IoT applications. However, prevailing TTA methods, which typically update all batch normalization (BN) layers, exhibit two memory inefficiencies. First, the reliance on BN layers for adaptation necessitates large batch sizes, leading to high memory usage. Second, updating all BN layers requires storing the activations of all BN layers for backpropagation, exacerbating the memory demand. Both factors lead to substantial memory costs, making existing solutions impractical for IoT devices. In this paper, we present FoCTTA, a low-memory CTTA strategy. The key is to automatically identify and adapt a few drift-sensitive representation layers, rather than blindly update all BN layers. The shift from BN to representation layers eliminates the need for large batch sizes. Also, by updating adaptation-critical layers only, FoCTTA avoids storing excessive activations. This focused adaptation approach ensures that FoCTTA is not only memory-efficient but also maintains effective adaptation. Evaluations show that FoCTTA improves the adaptation accuracy over the state-of-the-arts by 4.5%, 4.9%, and 14.8% on CIFAR10-C, CIFAR100-C, and ImageNet-C under the same memory constraints. Across various batch sizes, FoCTTA reduces the memory usage by 3-fold on average, while improving the accuracy by 8.1%, 3.6%, and 0.2%, respectively, on the three datasets.
no_new_dataset
0.948537
2502.20681
Zixuan Gong
Zixuan Gong, Jiaye Teng, Yong Liu
Disentangling Feature Structure: A Mathematically Provable Two-Stage Training Dynamics in Transformers
null
null
null
null
cs.CL cs.AI cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Transformers may exhibit two-stage training dynamics during the real-world training process. For instance, when training GPT-2 on the Counterfact dataset, the answers progress from syntactically incorrect to syntactically correct to semantically correct. However, existing theoretical analyses hardly account for this two-stage phenomenon. In this paper, we theoretically demonstrate how such two-stage training dynamics occur in transformers. Specifically, we analyze the dynamics of transformers using feature learning techniques under in-context learning regimes, based on a disentangled two-type feature structure. Such disentanglement of feature structure is general in practice, e.g., natural languages contain syntax and semantics, and proteins contain primary and secondary structures. To our best known, this is the first rigorous result regarding a two-stage optimization process in transformers. Additionally, a corollary indicates that such a two-stage process is closely related to the spectral properties of the attention weights, which accords well with empirical findings.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:27:24 GMT" } ]
2025-03-03T00:00:00
[ [ "Gong", "Zixuan", "" ], [ "Teng", "Jiaye", "" ], [ "Liu", "Yong", "" ] ]
TITLE: Disentangling Feature Structure: A Mathematically Provable Two-Stage Training Dynamics in Transformers ABSTRACT: Transformers may exhibit two-stage training dynamics during the real-world training process. For instance, when training GPT-2 on the Counterfact dataset, the answers progress from syntactically incorrect to syntactically correct to semantically correct. However, existing theoretical analyses hardly account for this two-stage phenomenon. In this paper, we theoretically demonstrate how such two-stage training dynamics occur in transformers. Specifically, we analyze the dynamics of transformers using feature learning techniques under in-context learning regimes, based on a disentangled two-type feature structure. Such disentanglement of feature structure is general in practice, e.g., natural languages contain syntax and semantics, and proteins contain primary and secondary structures. To our best known, this is the first rigorous result regarding a two-stage optimization process in transformers. Additionally, a corollary indicates that such a two-stage process is closely related to the spectral properties of the attention weights, which accords well with empirical findings.
no_new_dataset
0.947478
2502.20682
Gibson Nkhata
Gibson Nkhata, Susan Gauch, Usman Anjum and Justin Zhan
Fine-tuning BERT with Bidirectional LSTM for Fine-grained Movie Reviews Sentiment Analysis
14 pages, 5 figures, published in International Journal On Advances in Systems and Measurements, volume 16, numbers 3 and 4, 2023
null
null
null
cs.CL cs.AI
http://creativecommons.org/licenses/by-nc-nd/4.0/
Sentiment Analysis (SA) is instrumental in understanding peoples viewpoints facilitating social media monitoring recognizing products and brands and gauging customer satisfaction. Consequently SA has evolved into an active research domain within Natural Language Processing (NLP). Many approaches outlined in the literature devise intricate frameworks aimed at achieving high accuracy, focusing exclusively on either binary sentiment classification or fine-grained sentiment classification. In this paper our objective is to fine-tune the pre-trained BERT model with Bidirectional LSTM (BiLSTM) to enhance both binary and fine-grained SA specifically for movie reviews. Our approach involves conducting sentiment classification for each review followed by computing the overall sentiment polarity across all reviews. We present our findings on binary classification as well as fine-grained classification utilizing benchmark datasets. Additionally we implement and assess two accuracy improvement techniques Synthetic Minority Oversampling Technique (SMOTE) and NLP Augmenter (NLPAUG) to bolster the models generalization in fine-grained sentiment classification. Finally a heuristic algorithm is employed to calculate the overall polarity of predicted reviews from the BERT+BiLSTM output vector. Our approach performs comparably with state-of-the-art (SOTA) techniques in both classifications. For instance in binary classification we achieve 97.67% accuracy surpassing the leading SOTA model NB-weighted-BON+dv-cosine by 0.27% on the renowned IMDb dataset. Conversely for five-class classification on SST-5 while the top SOTA model RoBERTa+large+Self-explaining attains 55.5% accuracy our model achieves 59.48% accuracy surpassing the BERT-large baseline by 3.6%.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:30:48 GMT" } ]
2025-03-03T00:00:00
[ [ "Nkhata", "Gibson", "" ], [ "Gauch", "Susan", "" ], [ "Anjum", "Usman", "" ], [ "Zhan", "Justin", "" ] ]
TITLE: Fine-tuning BERT with Bidirectional LSTM for Fine-grained Movie Reviews Sentiment Analysis ABSTRACT: Sentiment Analysis (SA) is instrumental in understanding peoples viewpoints facilitating social media monitoring recognizing products and brands and gauging customer satisfaction. Consequently SA has evolved into an active research domain within Natural Language Processing (NLP). Many approaches outlined in the literature devise intricate frameworks aimed at achieving high accuracy, focusing exclusively on either binary sentiment classification or fine-grained sentiment classification. In this paper our objective is to fine-tune the pre-trained BERT model with Bidirectional LSTM (BiLSTM) to enhance both binary and fine-grained SA specifically for movie reviews. Our approach involves conducting sentiment classification for each review followed by computing the overall sentiment polarity across all reviews. We present our findings on binary classification as well as fine-grained classification utilizing benchmark datasets. Additionally we implement and assess two accuracy improvement techniques Synthetic Minority Oversampling Technique (SMOTE) and NLP Augmenter (NLPAUG) to bolster the models generalization in fine-grained sentiment classification. Finally a heuristic algorithm is employed to calculate the overall polarity of predicted reviews from the BERT+BiLSTM output vector. Our approach performs comparably with state-of-the-art (SOTA) techniques in both classifications. For instance in binary classification we achieve 97.67% accuracy surpassing the leading SOTA model NB-weighted-BON+dv-cosine by 0.27% on the renowned IMDb dataset. Conversely for five-class classification on SST-5 while the top SOTA model RoBERTa+large+Self-explaining attains 55.5% accuracy our model achieves 59.48% accuracy surpassing the BERT-large baseline by 3.6%.
no_new_dataset
0.953794
2502.20684
Abhishek Kumar Umrawal
Yingbing Huang, Deming Chen, and Abhishek K. Umrawal
JAM: Controllable and Responsible Text Generation via Causal Reasoning and Latent Vector Manipulation
10 pages, 3 figures, and 6 tables
null
null
null
cs.CL cs.AI cs.LG
http://creativecommons.org/licenses/by-nc-nd/4.0/
While large language models (LLMs) have made significant strides in generating coherent and contextually relevant text, they often function as opaque black boxes, trained on vast unlabeled datasets with statistical objectives, lacking an interpretable framework for responsible control. In this paper, we introduce JAM (Just A Move), a novel framework that interprets and controls text generation by integrating cause-effect analysis within the latent space of LLMs. Based on our observations, we uncover the inherent causality in LLM generation, which is critical for producing responsible and realistic outputs. Moreover, we explore latent vectors as fundamental components in LLM architectures, aiming to understand and manipulate them for more effective and efficient controllable text generation. We evaluate our framework using a range of tools, including the HHH criteria, toxicity reduction benchmarks, and GPT-4 alignment measures. Our results show that JAM achieves up to a 22% improvement over previous Controllable Text Generation (CTG) methods across multiple quantitative metrics and human-centric evaluations. Furthermore, JAM demonstrates greater computational efficiency compared to other CTG methods. These results highlight the effectiveness and efficiency of JAM for responsible and realistic text generation, paving the way for more interpretable and controllable models.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:31:48 GMT" } ]
2025-03-03T00:00:00
[ [ "Huang", "Yingbing", "" ], [ "Chen", "Deming", "" ], [ "Umrawal", "Abhishek K.", "" ] ]
TITLE: JAM: Controllable and Responsible Text Generation via Causal Reasoning and Latent Vector Manipulation ABSTRACT: While large language models (LLMs) have made significant strides in generating coherent and contextually relevant text, they often function as opaque black boxes, trained on vast unlabeled datasets with statistical objectives, lacking an interpretable framework for responsible control. In this paper, we introduce JAM (Just A Move), a novel framework that interprets and controls text generation by integrating cause-effect analysis within the latent space of LLMs. Based on our observations, we uncover the inherent causality in LLM generation, which is critical for producing responsible and realistic outputs. Moreover, we explore latent vectors as fundamental components in LLM architectures, aiming to understand and manipulate them for more effective and efficient controllable text generation. We evaluate our framework using a range of tools, including the HHH criteria, toxicity reduction benchmarks, and GPT-4 alignment measures. Our results show that JAM achieves up to a 22% improvement over previous Controllable Text Generation (CTG) methods across multiple quantitative metrics and human-centric evaluations. Furthermore, JAM demonstrates greater computational efficiency compared to other CTG methods. These results highlight the effectiveness and efficiency of JAM for responsible and realistic text generation, paving the way for more interpretable and controllable models.
no_new_dataset
0.945701
2502.20685
Dongki Jung
Dongki Jung, Jaehoon Choi, Yonghan Lee, Somi Jeong, Taejae Lee, Dinesh Manocha, Suyong Yeon
EDM: Equirectangular Projection-Oriented Dense Kernelized Feature Matching
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
We introduce the first learning-based dense matching algorithm, termed Equirectangular Projection-Oriented Dense Kernelized Feature Matching (EDM), specifically designed for omnidirectional images. Equirectangular projection (ERP) images, with their large fields of view, are particularly suited for dense matching techniques that aim to establish comprehensive correspondences across images. However, ERP images are subject to significant distortions, which we address by leveraging the spherical camera model and geodesic flow refinement in the dense matching method. To further mitigate these distortions, we propose spherical positional embeddings based on 3D Cartesian coordinates of the feature grid. Additionally, our method incorporates bidirectional transformations between spherical and Cartesian coordinate systems during refinement, utilizing a unit sphere to improve matching performance. We demonstrate that our proposed method achieves notable performance enhancements, with improvements of +26.72 and +42.62 in AUC@5{\deg} on the Matterport3D and Stanford2D3D datasets.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:37:01 GMT" } ]
2025-03-03T00:00:00
[ [ "Jung", "Dongki", "" ], [ "Choi", "Jaehoon", "" ], [ "Lee", "Yonghan", "" ], [ "Jeong", "Somi", "" ], [ "Lee", "Taejae", "" ], [ "Manocha", "Dinesh", "" ], [ "Yeon", "Suyong", "" ] ]
TITLE: EDM: Equirectangular Projection-Oriented Dense Kernelized Feature Matching ABSTRACT: We introduce the first learning-based dense matching algorithm, termed Equirectangular Projection-Oriented Dense Kernelized Feature Matching (EDM), specifically designed for omnidirectional images. Equirectangular projection (ERP) images, with their large fields of view, are particularly suited for dense matching techniques that aim to establish comprehensive correspondences across images. However, ERP images are subject to significant distortions, which we address by leveraging the spherical camera model and geodesic flow refinement in the dense matching method. To further mitigate these distortions, we propose spherical positional embeddings based on 3D Cartesian coordinates of the feature grid. Additionally, our method incorporates bidirectional transformations between spherical and Cartesian coordinate systems during refinement, utilizing a unit sphere to improve matching performance. We demonstrate that our proposed method achieves notable performance enhancements, with improvements of +26.72 and +42.62 in AUC@5{\deg} on the Matterport3D and Stanford2D3D datasets.
no_new_dataset
0.951459
2502.20687
Yihan Wang
Yihan Wang, Fei Xiong, Zhexin Han, Qi Song, Kaiqiao Zhan, Ben Wang
Unleashing the Potential of Two-Tower Models: Diffusion-Based Cross-Interaction for Large-Scale Matching
null
null
null
null
cs.IR cs.AI
http://creativecommons.org/licenses/by-sa/4.0/
Two-tower models are widely adopted in the industrial-scale matching stage across a broad range of application domains, such as content recommendations, advertisement systems, and search engines. This model efficiently handles large-scale candidate item screening by separating user and item representations. However, the decoupling network also leads to a neglect of potential information interaction between the user and item representations. Current state-of-the-art (SOTA) approaches include adding a shallow fully connected layer(i.e., COLD), which is limited by performance and can only be used in the ranking stage. For performance considerations, another approach attempts to capture historical positive interaction information from the other tower by regarding them as the input features(i.e., DAT). Later research showed that the gains achieved by this method are still limited because of lacking the guidance on the next user intent. To address the aforementioned challenges, we propose a "cross-interaction decoupling architecture" within our matching paradigm. This user-tower architecture leverages a diffusion module to reconstruct the next positive intention representation and employs a mixed-attention module to facilitate comprehensive cross-interaction. During the next positive intention generation, we further enhance the accuracy of its reconstruction by explicitly extracting the temporal drift within user behavior sequences. Experiments on two real-world datasets and one industrial dataset demonstrate that our method outperforms the SOTA two-tower models significantly, and our diffusion approach outperforms other generative models in reconstructing item representations.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 03:40:37 GMT" } ]
2025-03-03T00:00:00
[ [ "Wang", "Yihan", "" ], [ "Xiong", "Fei", "" ], [ "Han", "Zhexin", "" ], [ "Song", "Qi", "" ], [ "Zhan", "Kaiqiao", "" ], [ "Wang", "Ben", "" ] ]
TITLE: Unleashing the Potential of Two-Tower Models: Diffusion-Based Cross-Interaction for Large-Scale Matching ABSTRACT: Two-tower models are widely adopted in the industrial-scale matching stage across a broad range of application domains, such as content recommendations, advertisement systems, and search engines. This model efficiently handles large-scale candidate item screening by separating user and item representations. However, the decoupling network also leads to a neglect of potential information interaction between the user and item representations. Current state-of-the-art (SOTA) approaches include adding a shallow fully connected layer(i.e., COLD), which is limited by performance and can only be used in the ranking stage. For performance considerations, another approach attempts to capture historical positive interaction information from the other tower by regarding them as the input features(i.e., DAT). Later research showed that the gains achieved by this method are still limited because of lacking the guidance on the next user intent. To address the aforementioned challenges, we propose a "cross-interaction decoupling architecture" within our matching paradigm. This user-tower architecture leverages a diffusion module to reconstruct the next positive intention representation and employs a mixed-attention module to facilitate comprehensive cross-interaction. During the next positive intention generation, we further enhance the accuracy of its reconstruction by explicitly extracting the temporal drift within user behavior sequences. Experiments on two real-world datasets and one industrial dataset demonstrate that our method outperforms the SOTA two-tower models significantly, and our diffusion approach outperforms other generative models in reconstructing item representations.
no_new_dataset
0.948251
2502.20695
Yiping Sun
Yang Shi, Yiping Sun, Jiaolong Du, Xiaocheng Zhong, Zhiyong Wang, Yao Hu
Scalable Overload-Aware Graph-Based Index Construction for 10-Billion-Scale Vector Similarity Search
Accepted by WWW'25
null
null
null
cs.IR
http://creativecommons.org/licenses/by-sa/4.0/
Approximate Nearest Neighbor Search (ANNS) is essential for modern data-driven applications that require efficient retrieval of top-k results from massive vector databases. Although existing graph-based ANNS algorithms achieve a high recall rate on billion-scale datasets, their slow construction speed and limited scalability hinder their applicability to large-scale industrial scenarios. In this paper, we introduce SOGAIC, the first Scalable Overload-Aware Graph-Based ANNS Index Construction system tailored for ultra-large-scale vector databases: 1) We propose a dynamic data partitioning algorithm with overload constraints that adaptively introduces overlaps among subsets; 2) To enable efficient distributed subgraph construction, we employ a load-balancing task scheduling framework combined with an agglomerative merging strategy; 3) Extensive experiments on various datasets demonstrate a reduction of 47.3% in average construction time compared to existing methods. The proposed method has also been successfully deployed in a real-world industrial search engine, managing over 10 billion daily updated vectors and serving hundreds of millions of users.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 04:03:23 GMT" } ]
2025-03-03T00:00:00
[ [ "Shi", "Yang", "" ], [ "Sun", "Yiping", "" ], [ "Du", "Jiaolong", "" ], [ "Zhong", "Xiaocheng", "" ], [ "Wang", "Zhiyong", "" ], [ "Hu", "Yao", "" ] ]
TITLE: Scalable Overload-Aware Graph-Based Index Construction for 10-Billion-Scale Vector Similarity Search ABSTRACT: Approximate Nearest Neighbor Search (ANNS) is essential for modern data-driven applications that require efficient retrieval of top-k results from massive vector databases. Although existing graph-based ANNS algorithms achieve a high recall rate on billion-scale datasets, their slow construction speed and limited scalability hinder their applicability to large-scale industrial scenarios. In this paper, we introduce SOGAIC, the first Scalable Overload-Aware Graph-Based ANNS Index Construction system tailored for ultra-large-scale vector databases: 1) We propose a dynamic data partitioning algorithm with overload constraints that adaptively introduces overlaps among subsets; 2) To enable efficient distributed subgraph construction, we employ a load-balancing task scheduling framework combined with an agglomerative merging strategy; 3) Extensive experiments on various datasets demonstrate a reduction of 47.3% in average construction time compared to existing methods. The proposed method has also been successfully deployed in a real-world industrial search engine, managing over 10 billion daily updated vectors and serving hundreds of millions of users.
no_new_dataset
0.946399
2502.20708
Saminul Haque
John Duchi, Saminul Haque, Rohith Kuditipudi
A fast and slightly robust covariance estimator
39 pages
null
null
null
cs.DS
http://creativecommons.org/licenses/by/4.0/
Let $\mathcal{Z} = \{Z_1, \dots, Z_n\} \stackrel{\mathrm{i.i.d.}}{\sim} P \subset \mathbb{R}^d$ from a distribution $P$ with mean zero and covariance $\Sigma$. Given a dataset $\mathcal{X}$ such that $d_{\mathrm{ham}}(\mathcal{X}, \mathcal{Z}) \leq \varepsilon n$, we are interested in finding an efficient estimator $\widehat{\Sigma}$ that achieves $\mathrm{err}(\widehat{\Sigma}, \Sigma) := \|\Sigma^{-\frac{1}{2}}\widehat{\Sigma}\Sigma^{-\frac{1}{2}} - I\| _{\mathrm{op}} \leq 1/2$. We focus on the low contamination regime $\varepsilon = o(1/\sqrt{d}$). In this regime, prior work required either $\Omega(d^{3/2})$ samples or runtime that is exponential in $d$. We present an algorithm that, for subgaussian data, has near-linear sample complexity $n = \widetilde{\Omega}(d)$ and runtime $O((n+d)^{\omega + \frac{1}{2}})$, where $\omega$ is the matrix multiplication exponent. We also show that this algorithm works for heavy-tailed data with near-linear sample complexity, but in a smaller regime of $\varepsilon$. Concurrent to our work, Diakonikolas et al. [2024] give Sum-of-Squares estimators that achieve similar sample complexity but with large polynomial runtime.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 04:35:23 GMT" } ]
2025-03-03T00:00:00
[ [ "Duchi", "John", "" ], [ "Haque", "Saminul", "" ], [ "Kuditipudi", "Rohith", "" ] ]
TITLE: A fast and slightly robust covariance estimator ABSTRACT: Let $\mathcal{Z} = \{Z_1, \dots, Z_n\} \stackrel{\mathrm{i.i.d.}}{\sim} P \subset \mathbb{R}^d$ from a distribution $P$ with mean zero and covariance $\Sigma$. Given a dataset $\mathcal{X}$ such that $d_{\mathrm{ham}}(\mathcal{X}, \mathcal{Z}) \leq \varepsilon n$, we are interested in finding an efficient estimator $\widehat{\Sigma}$ that achieves $\mathrm{err}(\widehat{\Sigma}, \Sigma) := \|\Sigma^{-\frac{1}{2}}\widehat{\Sigma}\Sigma^{-\frac{1}{2}} - I\| _{\mathrm{op}} \leq 1/2$. We focus on the low contamination regime $\varepsilon = o(1/\sqrt{d}$). In this regime, prior work required either $\Omega(d^{3/2})$ samples or runtime that is exponential in $d$. We present an algorithm that, for subgaussian data, has near-linear sample complexity $n = \widetilde{\Omega}(d)$ and runtime $O((n+d)^{\omega + \frac{1}{2}})$, where $\omega$ is the matrix multiplication exponent. We also show that this algorithm works for heavy-tailed data with near-linear sample complexity, but in a smaller regime of $\varepsilon$. Concurrent to our work, Diakonikolas et al. [2024] give Sum-of-Squares estimators that achieve similar sample complexity but with large polynomial runtime.
no_new_dataset
0.934395
2502.20709
Zhengyi Zhong
Zhengyi Zhong, Weidong Bao, Ji Wang, Shuai Zhang, Jingxuan Zhou, Lingjuan Lyu, Wei Yang Bryan Lim
Unlearning through Knowledge Overwriting: Reversible Federated Unlearning via Selective Sparse Adapter
Accepted by CVPR2025
null
null
null
cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Federated Learning is a promising paradigm for privacy-preserving collaborative model training. In practice, it is essential not only to continuously train the model to acquire new knowledge but also to guarantee old knowledge the right to be forgotten (i.e., federated unlearning), especially for privacy-sensitive information or harmful knowledge. However, current federated unlearning methods face several challenges, including indiscriminate unlearning of cross-client knowledge, irreversibility of unlearning, and significant unlearning costs. To this end, we propose a method named FUSED, which first identifies critical layers by analyzing each layer's sensitivity to knowledge and constructs sparse unlearning adapters for sensitive ones. Then, the adapters are trained without altering the original parameters, overwriting the unlearning knowledge with the remaining knowledge. This knowledge overwriting process enables FUSED to mitigate the effects of indiscriminate unlearning. Moreover, the introduction of independent adapters makes unlearning reversible and significantly reduces the unlearning costs. Finally, extensive experiments on three datasets across various unlearning scenarios demonstrate that FUSED's effectiveness is comparable to Retraining, surpassing all other baselines while greatly reducing unlearning costs.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 04:35:26 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhong", "Zhengyi", "" ], [ "Bao", "Weidong", "" ], [ "Wang", "Ji", "" ], [ "Zhang", "Shuai", "" ], [ "Zhou", "Jingxuan", "" ], [ "Lyu", "Lingjuan", "" ], [ "Lim", "Wei Yang Bryan", "" ] ]
TITLE: Unlearning through Knowledge Overwriting: Reversible Federated Unlearning via Selective Sparse Adapter ABSTRACT: Federated Learning is a promising paradigm for privacy-preserving collaborative model training. In practice, it is essential not only to continuously train the model to acquire new knowledge but also to guarantee old knowledge the right to be forgotten (i.e., federated unlearning), especially for privacy-sensitive information or harmful knowledge. However, current federated unlearning methods face several challenges, including indiscriminate unlearning of cross-client knowledge, irreversibility of unlearning, and significant unlearning costs. To this end, we propose a method named FUSED, which first identifies critical layers by analyzing each layer's sensitivity to knowledge and constructs sparse unlearning adapters for sensitive ones. Then, the adapters are trained without altering the original parameters, overwriting the unlearning knowledge with the remaining knowledge. This knowledge overwriting process enables FUSED to mitigate the effects of indiscriminate unlearning. Moreover, the introduction of independent adapters makes unlearning reversible and significantly reduces the unlearning costs. Finally, extensive experiments on three datasets across various unlearning scenarios demonstrate that FUSED's effectiveness is comparable to Retraining, surpassing all other baselines while greatly reducing unlearning costs.
no_new_dataset
0.948965
2502.20715
Kiranmayee Janardhan
Kiranmayee Janardhan, Christy Bobby Thomas
Glioma Classification using Multi-sequence MRI and Novel Wavelets-based Feature Fusion
18 pages, 11 figures, 6 tables, journal paper
Journal of Computational Analysis and Applications, 2024
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Glioma, a prevalent and heterogeneous tumor originating from the glial cells, can be differentiated as Low Grade Glioma (LGG) and High Grade Glioma (HGG) according to World Health Organization's norms. Classifying gliomas is essential for treatment protocols that depend extensively on subtype differentiation. For non-invasive glioma evaluation, Magnetic Resonance Imaging (MRI) offers vital information about the morphology and location of the the tumor. The versatility of MRI allows the classification of gliomas as LGG and HGG based on their texture, perfusion, and diffusion characteristics, and further for improving the diagnosis and providing tailored treatments. Nevertheless, the precise classification is complicated by tumor heterogeneity and overlapping radiomic characteristics. Thus, in this work, wavelet based novel fusion algorithm were implemented on multi-sequence T1, T1-contrast enhanced (T1CE), T2 and Fluid Attenuated Inversion Recovery (FLAIR) MRI images to compute the radiomics features. Furthermore, principal component analysis is applied to reduce the feature space and XGBoost, Support Vector Machine, and Random Forest Classifier are used for the classification. The result shows that the SVM algorithm performs comparatively well with an accuracy of 90.17%, precision of 91.04% and recall of 96.19%, F1-score of 93.53%, and AUC of 94.60% when implemented on BraTS 2018 dataset and with an accuracy of 91.34%, precision of 93.05% and recall of 96.13%, F1-score of 94.53%, and AUC of 93.71% for BraTS 2018 dataset. Thus, the proposed algorithm could be potentially implemented for the computer-aided diagnosis and grading system for gliomas.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 04:58:41 GMT" } ]
2025-03-03T00:00:00
[ [ "Janardhan", "Kiranmayee", "" ], [ "Thomas", "Christy Bobby", "" ] ]
TITLE: Glioma Classification using Multi-sequence MRI and Novel Wavelets-based Feature Fusion ABSTRACT: Glioma, a prevalent and heterogeneous tumor originating from the glial cells, can be differentiated as Low Grade Glioma (LGG) and High Grade Glioma (HGG) according to World Health Organization's norms. Classifying gliomas is essential for treatment protocols that depend extensively on subtype differentiation. For non-invasive glioma evaluation, Magnetic Resonance Imaging (MRI) offers vital information about the morphology and location of the the tumor. The versatility of MRI allows the classification of gliomas as LGG and HGG based on their texture, perfusion, and diffusion characteristics, and further for improving the diagnosis and providing tailored treatments. Nevertheless, the precise classification is complicated by tumor heterogeneity and overlapping radiomic characteristics. Thus, in this work, wavelet based novel fusion algorithm were implemented on multi-sequence T1, T1-contrast enhanced (T1CE), T2 and Fluid Attenuated Inversion Recovery (FLAIR) MRI images to compute the radiomics features. Furthermore, principal component analysis is applied to reduce the feature space and XGBoost, Support Vector Machine, and Random Forest Classifier are used for the classification. The result shows that the SVM algorithm performs comparatively well with an accuracy of 90.17%, precision of 91.04% and recall of 96.19%, F1-score of 93.53%, and AUC of 94.60% when implemented on BraTS 2018 dataset and with an accuracy of 91.34%, precision of 93.05% and recall of 96.13%, F1-score of 94.53%, and AUC of 93.71% for BraTS 2018 dataset. Thus, the proposed algorithm could be potentially implemented for the computer-aided diagnosis and grading system for gliomas.
no_new_dataset
0.956472
2502.20719
Guanglin Zhou
Guanglin Zhou and Sebastiano Barbieri
Generating Clinically Realistic EHR Data via a Hierarchy- and Semantics-Guided Transformer
null
null
null
null
cs.LG cs.AI
http://creativecommons.org/licenses/by-nc-sa/4.0/
Generating realistic synthetic electronic health records (EHRs) holds tremendous promise for accelerating healthcare research, facilitating AI model development and enhancing patient privacy. However, existing generative methods typically treat EHRs as flat sequences of discrete medical codes. This approach overlooks two critical aspects: the inherent hierarchical organization of clinical coding systems and the rich semantic context provided by code descriptions. Consequently, synthetic patient sequences often lack high clinical fidelity and have limited utility in downstream clinical tasks. In this paper, we propose the Hierarchy- and Semantics-Guided Transformer (HiSGT), a novel framework that leverages both hierarchical and semantic information for the generative process. HiSGT constructs a hierarchical graph to encode parent-child and sibling relationships among clinical codes and employs a graph neural network to derive hierarchy-aware embeddings. These are then fused with semantic embeddings extracted from a pre-trained clinical language model (e.g., ClinicalBERT), enabling the Transformer-based generator to more accurately model the nuanced clinical patterns inherent in real EHRs. Extensive experiments on the MIMIC-III and MIMIC-IV datasets demonstrate that HiSGT significantly improves the statistical alignment of synthetic data with real patient records, as well as supports robust downstream applications such as chronic disease classification. By addressing the limitations of conventional raw code-based generative models, HiSGT represents a significant step toward clinically high-fidelity synthetic data generation and a general paradigm suitable for interpretable medical code representation, offering valuable applications in data augmentation and privacy-preserving healthcare analytics.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 05:06:04 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhou", "Guanglin", "" ], [ "Barbieri", "Sebastiano", "" ] ]
TITLE: Generating Clinically Realistic EHR Data via a Hierarchy- and Semantics-Guided Transformer ABSTRACT: Generating realistic synthetic electronic health records (EHRs) holds tremendous promise for accelerating healthcare research, facilitating AI model development and enhancing patient privacy. However, existing generative methods typically treat EHRs as flat sequences of discrete medical codes. This approach overlooks two critical aspects: the inherent hierarchical organization of clinical coding systems and the rich semantic context provided by code descriptions. Consequently, synthetic patient sequences often lack high clinical fidelity and have limited utility in downstream clinical tasks. In this paper, we propose the Hierarchy- and Semantics-Guided Transformer (HiSGT), a novel framework that leverages both hierarchical and semantic information for the generative process. HiSGT constructs a hierarchical graph to encode parent-child and sibling relationships among clinical codes and employs a graph neural network to derive hierarchy-aware embeddings. These are then fused with semantic embeddings extracted from a pre-trained clinical language model (e.g., ClinicalBERT), enabling the Transformer-based generator to more accurately model the nuanced clinical patterns inherent in real EHRs. Extensive experiments on the MIMIC-III and MIMIC-IV datasets demonstrate that HiSGT significantly improves the statistical alignment of synthetic data with real patient records, as well as supports robust downstream applications such as chronic disease classification. By addressing the limitations of conventional raw code-based generative models, HiSGT represents a significant step toward clinically high-fidelity synthetic data generation and a general paradigm suitable for interpretable medical code representation, offering valuable applications in data augmentation and privacy-preserving healthcare analytics.
no_new_dataset
0.948585
2502.20729
Mostafa Rahimi Azghadi
Ben Walters, Yeshwanth Bethi, Taylor Kergan, Binh Nguyen, Amirali Amirsoleimani, Jason K. Eshraghian, Saeed Afshar, Mostafa Rahimi Azghadi
NeuroMorse: A Temporally Structured Dataset For Neuromorphic Computing
null
null
null
null
cs.NE cs.AI
http://creativecommons.org/licenses/by/4.0/
Neuromorphic engineering aims to advance computing by mimicking the brain's efficient processing, where data is encoded as asynchronous temporal events. This eliminates the need for a synchronisation clock and minimises power consumption when no data is present. However, many benchmarks for neuromorphic algorithms primarily focus on spatial features, neglecting the temporal dynamics that are inherent to most sequence-based tasks. This gap may lead to evaluations that fail to fully capture the unique strengths and characteristics of neuromorphic systems. In this paper, we present NeuroMorse, a temporally structured dataset designed for benchmarking neuromorphic learning systems. NeuroMorse converts the top 50 words in the English language into temporal Morse code spike sequences. Despite using only two input spike channels for Morse dots and dashes, complex information is encoded through temporal patterns in the data. The proposed benchmark contains feature hierarchy at multiple temporal scales that test the capacity of neuromorphic algorithms to decompose input patterns into spatial and temporal hierarchies. We demonstrate that our training set is challenging to categorise using a linear classifier and that identifying keywords in the test set is difficult using conventional methods. The NeuroMorse dataset is available at Zenodo, with our accompanying code on GitHub at https://github.com/Ben-E-Walters/NeuroMorse.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 05:22:45 GMT" } ]
2025-03-03T00:00:00
[ [ "Walters", "Ben", "" ], [ "Bethi", "Yeshwanth", "" ], [ "Kergan", "Taylor", "" ], [ "Nguyen", "Binh", "" ], [ "Amirsoleimani", "Amirali", "" ], [ "Eshraghian", "Jason K.", "" ], [ "Afshar", "Saeed", "" ], [ "Azghadi", "Mostafa Rahimi", "" ] ]
TITLE: NeuroMorse: A Temporally Structured Dataset For Neuromorphic Computing ABSTRACT: Neuromorphic engineering aims to advance computing by mimicking the brain's efficient processing, where data is encoded as asynchronous temporal events. This eliminates the need for a synchronisation clock and minimises power consumption when no data is present. However, many benchmarks for neuromorphic algorithms primarily focus on spatial features, neglecting the temporal dynamics that are inherent to most sequence-based tasks. This gap may lead to evaluations that fail to fully capture the unique strengths and characteristics of neuromorphic systems. In this paper, we present NeuroMorse, a temporally structured dataset designed for benchmarking neuromorphic learning systems. NeuroMorse converts the top 50 words in the English language into temporal Morse code spike sequences. Despite using only two input spike channels for Morse dots and dashes, complex information is encoded through temporal patterns in the data. The proposed benchmark contains feature hierarchy at multiple temporal scales that test the capacity of neuromorphic algorithms to decompose input patterns into spatial and temporal hierarchies. We demonstrate that our training set is challenging to categorise using a linear classifier and that identifying keywords in the test set is difficult using conventional methods. The NeuroMorse dataset is available at Zenodo, with our accompanying code on GitHub at https://github.com/Ben-E-Walters/NeuroMorse.
new_dataset
0.957755
2502.20731
Sean Kouma
Sean Kouma, Rachel Masters
Indoor Localization for Autonomous Robot Navigation
10 pages, 6 figures
null
null
null
cs.RO cs.LG
http://creativecommons.org/licenses/by/4.0/
Indoor positioning systems (IPSs) have gained attention as outdoor navigation becomes prevalent in everyday life. Research is being actively conducted on how indoor smartphone navigation can be accomplished and improved using received signal strength indication (RSSI) and machine learning (ML). IPSs have more use cases that need further exploration, and we aim to explore using IPSs for the indoor navigation of an autonomous robot. We collected a dataset and trained models to test on a robot. We also developed an A* path-planning algorithm so that our robot could navigate itself using predicted directions. After testing different network structures, our robot was able to successfully navigate corners around 50 percent of the time. The findings of this paper indicate that using IPSs for autonomous robots is a promising area of future research.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 05:25:04 GMT" } ]
2025-03-03T00:00:00
[ [ "Kouma", "Sean", "" ], [ "Masters", "Rachel", "" ] ]
TITLE: Indoor Localization for Autonomous Robot Navigation ABSTRACT: Indoor positioning systems (IPSs) have gained attention as outdoor navigation becomes prevalent in everyday life. Research is being actively conducted on how indoor smartphone navigation can be accomplished and improved using received signal strength indication (RSSI) and machine learning (ML). IPSs have more use cases that need further exploration, and we aim to explore using IPSs for the indoor navigation of an autonomous robot. We collected a dataset and trained models to test on a robot. We also developed an A* path-planning algorithm so that our robot could navigate itself using predicted directions. After testing different network structures, our robot was able to successfully navigate corners around 50 percent of the time. The findings of this paper indicate that using IPSs for autonomous robots is a promising area of future research.
new_dataset
0.960025
2502.20749
Yichi Zhang
Yichi Zhang, Bohao Lv, Le Xue, Wenbo Zhang, Yuchen Liu, Yu Fu, Yuan Cheng, Yuan Qi
SemiSAM+: Rethinking Semi-Supervised Medical Image Segmentation in the Era of Foundation Models
null
null
null
null
eess.IV cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Deep learning-based medical image segmentation typically requires large amount of labeled data for training, making it less applicable in clinical settings due to high annotation cost. Semi-supervised learning (SSL) has emerged as an appealing strategy due to its less dependence on acquiring abundant annotations from experts compared to fully supervised methods. Beyond existing model-centric advancements of SSL by designing novel regularization strategies, we anticipate a paradigmatic shift due to the emergence of promptable segmentation foundation models with universal segmentation capabilities using positional prompts represented by Segment Anything Model (SAM). In this paper, we present SemiSAM+, a foundation model-driven SSL framework to efficiently learn from limited labeled data for medical image segmentation. SemiSAM+ consists of one or multiple promptable foundation models as generalist models, and a trainable task-specific segmentation model as specialist model. For a given new segmentation task, the training is based on the specialist-generalist collaborative learning procedure, where the trainable specialist model delivers positional prompts to interact with the frozen generalist models to acquire pseudo-labels, and then the generalist model output provides the specialist model with informative and efficient supervision which benefits the automatic segmentation and prompt generation in turn. Extensive experiments on two public datasets and one in-house clinical dataset demonstrate that SemiSAM+ achieves significant performance improvement, especially under extremely limited annotation scenarios, and shows strong efficiency as a plug-and-play strategy that can be easily adapted to different specialist and generalist models.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 05:54:41 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhang", "Yichi", "" ], [ "Lv", "Bohao", "" ], [ "Xue", "Le", "" ], [ "Zhang", "Wenbo", "" ], [ "Liu", "Yuchen", "" ], [ "Fu", "Yu", "" ], [ "Cheng", "Yuan", "" ], [ "Qi", "Yuan", "" ] ]
TITLE: SemiSAM+: Rethinking Semi-Supervised Medical Image Segmentation in the Era of Foundation Models ABSTRACT: Deep learning-based medical image segmentation typically requires large amount of labeled data for training, making it less applicable in clinical settings due to high annotation cost. Semi-supervised learning (SSL) has emerged as an appealing strategy due to its less dependence on acquiring abundant annotations from experts compared to fully supervised methods. Beyond existing model-centric advancements of SSL by designing novel regularization strategies, we anticipate a paradigmatic shift due to the emergence of promptable segmentation foundation models with universal segmentation capabilities using positional prompts represented by Segment Anything Model (SAM). In this paper, we present SemiSAM+, a foundation model-driven SSL framework to efficiently learn from limited labeled data for medical image segmentation. SemiSAM+ consists of one or multiple promptable foundation models as generalist models, and a trainable task-specific segmentation model as specialist model. For a given new segmentation task, the training is based on the specialist-generalist collaborative learning procedure, where the trainable specialist model delivers positional prompts to interact with the frozen generalist models to acquire pseudo-labels, and then the generalist model output provides the specialist model with informative and efficient supervision which benefits the automatic segmentation and prompt generation in turn. Extensive experiments on two public datasets and one in-house clinical dataset demonstrate that SemiSAM+ achieves significant performance improvement, especially under extremely limited annotation scenarios, and shows strong efficiency as a plug-and-play strategy that can be easily adapted to different specialist and generalist models.
no_new_dataset
0.951639
2502.20755
Soumya Mukherjee
Soumya Mukherjee, Bharath K. Sriperumbudur
Minimax Optimal Kernel Two-Sample Tests with Random Features
82 pages, 10 figures, 5 tables
null
null
null
math.ST cs.LG stat.ML stat.TH
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Reproducing Kernel Hilbert Space (RKHS) embedding of probability distributions has proved to be an effective approach, via MMD (maximum mean discrepancy) for nonparametric hypothesis testing problems involving distributions defined over general (non-Euclidean) domains. While a substantial amount of work has been done on this topic, only recently, minimax optimal two-sample tests have been constructed that incorporate, unlike MMD, both the mean element and a regularized version of the covariance operator. However, as with most kernel algorithms, the computational complexity of the optimal test scales cubically in the sample size, limiting its applicability. In this paper, we propose a spectral regularized two-sample test based on random Fourier feature (RFF) approximation and investigate the trade-offs between statistical optimality and computational efficiency. We show the proposed test to be minimax optimal if the approximation order of RFF (which depends on the smoothness of the likelihood ratio and the decay rate of the eigenvalues of the integral operator) is sufficiently large. We develop a practically implementable permutation-based version of the proposed test with a data-adaptive strategy for selecting the regularization parameter and the kernel. Finally, through numerical experiments on simulated and benchmark datasets, we demonstrate that the proposed RFF-based test is computationally efficient and performs almost similar (with a small drop in power) to the exact test.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 06:12:00 GMT" } ]
2025-03-03T00:00:00
[ [ "Mukherjee", "Soumya", "" ], [ "Sriperumbudur", "Bharath K.", "" ] ]
TITLE: Minimax Optimal Kernel Two-Sample Tests with Random Features ABSTRACT: Reproducing Kernel Hilbert Space (RKHS) embedding of probability distributions has proved to be an effective approach, via MMD (maximum mean discrepancy) for nonparametric hypothesis testing problems involving distributions defined over general (non-Euclidean) domains. While a substantial amount of work has been done on this topic, only recently, minimax optimal two-sample tests have been constructed that incorporate, unlike MMD, both the mean element and a regularized version of the covariance operator. However, as with most kernel algorithms, the computational complexity of the optimal test scales cubically in the sample size, limiting its applicability. In this paper, we propose a spectral regularized two-sample test based on random Fourier feature (RFF) approximation and investigate the trade-offs between statistical optimality and computational efficiency. We show the proposed test to be minimax optimal if the approximation order of RFF (which depends on the smoothness of the likelihood ratio and the decay rate of the eigenvalues of the integral operator) is sufficiently large. We develop a practically implementable permutation-based version of the proposed test with a data-adaptive strategy for selecting the regularization parameter and the kernel. Finally, through numerical experiments on simulated and benchmark datasets, we demonstrate that the proposed RFF-based test is computationally efficient and performs almost similar (with a small drop in power) to the exact test.
no_new_dataset
0.944791
2502.20767
Hui Lai
Hui Lai, Qi Chen, Junping Zhang, Jian Pu
A2DO: Adaptive Anti-Degradation Odometry with Deep Multi-Sensor Fusion for Autonomous Navigation
6+1pages, 6 figures, accept by ICRA
null
null
null
cs.RO
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Accurate localization is essential for the safe and effective navigation of autonomous vehicles, and Simultaneous Localization and Mapping (SLAM) is a cornerstone technology in this context. However, The performance of the SLAM system can deteriorate under challenging conditions such as low light, adverse weather, or obstructions due to sensor degradation. We present A2DO, a novel end-to-end multi-sensor fusion odometry system that enhances robustness in these scenarios through deep neural networks. A2DO integrates LiDAR and visual data, employing a multi-layer, multi-scale feature encoding module augmented by an attention mechanism to mitigate sensor degradation dynamically. The system is pre-trained extensively on simulated datasets covering a broad range of degradation scenarios and fine-tuned on a curated set of real-world data, ensuring robust adaptation to complex scenarios. Our experiments demonstrate that A2DO maintains superior localization accuracy and robustness across various degradation conditions, showcasing its potential for practical implementation in autonomous vehicle systems.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 06:37:51 GMT" } ]
2025-03-03T00:00:00
[ [ "Lai", "Hui", "" ], [ "Chen", "Qi", "" ], [ "Zhang", "Junping", "" ], [ "Pu", "Jian", "" ] ]
TITLE: A2DO: Adaptive Anti-Degradation Odometry with Deep Multi-Sensor Fusion for Autonomous Navigation ABSTRACT: Accurate localization is essential for the safe and effective navigation of autonomous vehicles, and Simultaneous Localization and Mapping (SLAM) is a cornerstone technology in this context. However, The performance of the SLAM system can deteriorate under challenging conditions such as low light, adverse weather, or obstructions due to sensor degradation. We present A2DO, a novel end-to-end multi-sensor fusion odometry system that enhances robustness in these scenarios through deep neural networks. A2DO integrates LiDAR and visual data, employing a multi-layer, multi-scale feature encoding module augmented by an attention mechanism to mitigate sensor degradation dynamically. The system is pre-trained extensively on simulated datasets covering a broad range of degradation scenarios and fine-tuned on a curated set of real-world data, ensuring robust adaptation to complex scenarios. Our experiments demonstrate that A2DO maintains superior localization accuracy and robustness across various degradation conditions, showcasing its potential for practical implementation in autonomous vehicle systems.
no_new_dataset
0.941547
2502.20772
Tianyi Wang
Tianyi Zeng, Tianyi Wang, Junfeng Jiao, Xinbo Chen
Damper-B-PINN: Damper Characteristics-Based Bayesian Physics-Informed Neural Network for Vehicle State Estimation
null
null
null
null
cs.AI cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
State estimation for Multi-Input Multi-Output (MIMO) systems with noise, such as vehicle chassis systems, presents a significant challenge due to the imperfect and complex relationship between inputs and outputs. To solve this problem, we design a Damper characteristics-based Bayesian Physics-Informed Neural Network (Damper-B-PINN). First, we introduce a neuron forward process inspired by the mechanical properties of dampers, which limits abrupt jumps in neuron values between epochs while maintaining search capability. Additionally, we apply an optimized Bayesian dropout layer to the MIMO system to enhance robustness against noise and prevent non-convergence issues. Physical information is incorporated into the loss function to serve as a physical prior for the neural network. The effectiveness of our Damper-B-PINN architecture is then validated across ten datasets and fourteen vehicle types, demonstrating superior accuracy, computational efficiency, and convergence in vehicle state estimation (i.e., dynamic wheel load) compared to other state-of-the-art benchmarks.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 06:46:21 GMT" } ]
2025-03-03T00:00:00
[ [ "Zeng", "Tianyi", "" ], [ "Wang", "Tianyi", "" ], [ "Jiao", "Junfeng", "" ], [ "Chen", "Xinbo", "" ] ]
TITLE: Damper-B-PINN: Damper Characteristics-Based Bayesian Physics-Informed Neural Network for Vehicle State Estimation ABSTRACT: State estimation for Multi-Input Multi-Output (MIMO) systems with noise, such as vehicle chassis systems, presents a significant challenge due to the imperfect and complex relationship between inputs and outputs. To solve this problem, we design a Damper characteristics-based Bayesian Physics-Informed Neural Network (Damper-B-PINN). First, we introduce a neuron forward process inspired by the mechanical properties of dampers, which limits abrupt jumps in neuron values between epochs while maintaining search capability. Additionally, we apply an optimized Bayesian dropout layer to the MIMO system to enhance robustness against noise and prevent non-convergence issues. Physical information is incorporated into the loss function to serve as a physical prior for the neural network. The effectiveness of our Damper-B-PINN architecture is then validated across ten datasets and fourteen vehicle types, demonstrating superior accuracy, computational efficiency, and convergence in vehicle state estimation (i.e., dynamic wheel load) compared to other state-of-the-art benchmarks.
no_new_dataset
0.950824
2502.20777
Jianping Yao
Long Huang and Jianping Yao
Silicon Micro-Disk Resonator Crossbar Array for High-Speed and High-Density Photonic Convolution Processing
15 pages, 23 figures
null
null
null
physics.optics
http://creativecommons.org/licenses/by-nc-nd/4.0/
Advanced artificial intelligence (AI) algorithms, particularly those based on artificial neural networks, have garnered significant attention for their potential applications in areas such as image recognition and natural language processing. Notably, neural networks make heavy use of matrix-vector multiplication (MVM) operations, causing substantial computing burden on existing electronic computing systems. Optical computing has attracted considerable attention that can perform optical-domain MVM at an ultra-high speed. In this paper, we introduce a novel silicon photonic micro-disk resonator (MDR) crossbar signal processor designed to support matrix-vector multiplication (MVM) with both high processing speed and enhanced computational density. The key innovation of the proposed MDR crossbar processor is the placement of two MDRs at each crosspoint, enabling simultaneous routing and weighting functions. This design effectively doubles the computational density, improving overall performance. We fabricate a silicon photonic MDR crossbar processor, which is employed to perform convolutional tasks in a convolutional neural network (CNN). The experimental results demonstrate that the photonic processor achieves a classification accuracy of 96% on the MNIST dataset. Additionally, it is capable of scaling to a computational speed of up to 160 tera-operations per second (TOPS) and a computational density as high as 25.6 TOPS/mm2. Our approach holds significant promise for enabling highly efficient, scalable on-chip optical computing, with broad potential applications in AI and beyond.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 06:55:45 GMT" } ]
2025-03-03T00:00:00
[ [ "Huang", "Long", "" ], [ "Yao", "Jianping", "" ] ]
TITLE: Silicon Micro-Disk Resonator Crossbar Array for High-Speed and High-Density Photonic Convolution Processing ABSTRACT: Advanced artificial intelligence (AI) algorithms, particularly those based on artificial neural networks, have garnered significant attention for their potential applications in areas such as image recognition and natural language processing. Notably, neural networks make heavy use of matrix-vector multiplication (MVM) operations, causing substantial computing burden on existing electronic computing systems. Optical computing has attracted considerable attention that can perform optical-domain MVM at an ultra-high speed. In this paper, we introduce a novel silicon photonic micro-disk resonator (MDR) crossbar signal processor designed to support matrix-vector multiplication (MVM) with both high processing speed and enhanced computational density. The key innovation of the proposed MDR crossbar processor is the placement of two MDRs at each crosspoint, enabling simultaneous routing and weighting functions. This design effectively doubles the computational density, improving overall performance. We fabricate a silicon photonic MDR crossbar processor, which is employed to perform convolutional tasks in a convolutional neural network (CNN). The experimental results demonstrate that the photonic processor achieves a classification accuracy of 96% on the MNIST dataset. Additionally, it is capable of scaling to a computational speed of up to 160 tera-operations per second (TOPS) and a computational density as high as 25.6 TOPS/mm2. Our approach holds significant promise for enabling highly efficient, scalable on-chip optical computing, with broad potential applications in AI and beyond.
no_new_dataset
0.953144
2502.20780
Qiao Yan
Qiao Yan, Yuchen Yuan, Xiaowei Hu, Yihan Wang, Jiaqi Xu, Jinpeng Li, Chi-Wing Fu, Pheng-Ann Heng
MedHallTune: An Instruction-Tuning Benchmark for Mitigating Medical Hallucination in Vision-Language Models
null
null
null
null
cs.AI cs.CL cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The increasing use of vision-language models (VLMs) in healthcare applications presents great challenges related to hallucinations, in which the models may generate seemingly plausible results that are in fact incorrect. Such hallucinations can jeopardize clinical decision making, potentially harming the diagnosis and treatments. In this work, we propose MedHallTune, a large-scale benchmark designed specifically to evaluate and mitigate hallucinations in medical VLMs. Comprising over 100,000 images and 1,000,000 instruction pairs, MedHallTune includes both hallucination and non-hallucination samples, each with ground-truth annotations. We conduct a comprehensive evaluation of current medical and general VLMs using MedHallTune, assessing their performance across key metrics, including clinical accuracy, relevance, detail level, and risk level. The experimental results show that fine-tuning with MedHallTune successfully improves the ability of several existing models to manage hallucinations and boost their zero-shot performance on downstream visual-question-answering (VQA) tasks, making them more reliable for practical medical applications. Our work contributes to the development of more trustworthy VLMs. Codes and dataset will be available at \href{https://github.com/russellyq/MedHallTune}{MedHallTune}.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 06:59:49 GMT" } ]
2025-03-03T00:00:00
[ [ "Yan", "Qiao", "" ], [ "Yuan", "Yuchen", "" ], [ "Hu", "Xiaowei", "" ], [ "Wang", "Yihan", "" ], [ "Xu", "Jiaqi", "" ], [ "Li", "Jinpeng", "" ], [ "Fu", "Chi-Wing", "" ], [ "Heng", "Pheng-Ann", "" ] ]
TITLE: MedHallTune: An Instruction-Tuning Benchmark for Mitigating Medical Hallucination in Vision-Language Models ABSTRACT: The increasing use of vision-language models (VLMs) in healthcare applications presents great challenges related to hallucinations, in which the models may generate seemingly plausible results that are in fact incorrect. Such hallucinations can jeopardize clinical decision making, potentially harming the diagnosis and treatments. In this work, we propose MedHallTune, a large-scale benchmark designed specifically to evaluate and mitigate hallucinations in medical VLMs. Comprising over 100,000 images and 1,000,000 instruction pairs, MedHallTune includes both hallucination and non-hallucination samples, each with ground-truth annotations. We conduct a comprehensive evaluation of current medical and general VLMs using MedHallTune, assessing their performance across key metrics, including clinical accuracy, relevance, detail level, and risk level. The experimental results show that fine-tuning with MedHallTune successfully improves the ability of several existing models to manage hallucinations and boost their zero-shot performance on downstream visual-question-answering (VQA) tasks, making them more reliable for practical medical applications. Our work contributes to the development of more trustworthy VLMs. Codes and dataset will be available at \href{https://github.com/russellyq/MedHallTune}{MedHallTune}.
new_dataset
0.962568
2502.20784
Tao Chen
Tao Chen and Chenhui Wang and Zhihao Chen and Hongming Shan
Autoregressive Medical Image Segmentation via Next-Scale Mask Prediction
10 pages, 4 figures
null
null
null
eess.IV cs.CV
http://creativecommons.org/licenses/by-nc-nd/4.0/
While deep learning has significantly advanced medical image segmentation, most existing methods still struggle with handling complex anatomical regions. Cascaded or deep supervision-based approaches attempt to address this challenge through multi-scale feature learning but fail to establish sufficient inter-scale dependencies, as each scale relies solely on the features of the immediate predecessor. To this end, we propose the AutoRegressive Segmentation framework via next-scale mask prediction, termed AR-Seg, which progressively predicts the next-scale mask by explicitly modeling dependencies across all previous scales within a unified architecture. AR-Seg introduces three innovations: (1) a multi-scale mask autoencoder that quantizes the mask into multi-scale token maps to capture hierarchical anatomical structures, (2) a next-scale autoregressive mechanism that progressively predicts next-scale masks to enable sufficient inter-scale dependencies, and (3) a consensus-aggregation strategy that combines multiple sampled results to generate a more accurate mask, further improving segmentation robustness. Extensive experimental results on two benchmark datasets with different modalities demonstrate that AR-Seg outperforms state-of-the-art methods while explicitly visualizing the intermediate coarse-to-fine segmentation process.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 07:05:58 GMT" } ]
2025-03-03T00:00:00
[ [ "Chen", "Tao", "" ], [ "Wang", "Chenhui", "" ], [ "Chen", "Zhihao", "" ], [ "Shan", "Hongming", "" ] ]
TITLE: Autoregressive Medical Image Segmentation via Next-Scale Mask Prediction ABSTRACT: While deep learning has significantly advanced medical image segmentation, most existing methods still struggle with handling complex anatomical regions. Cascaded or deep supervision-based approaches attempt to address this challenge through multi-scale feature learning but fail to establish sufficient inter-scale dependencies, as each scale relies solely on the features of the immediate predecessor. To this end, we propose the AutoRegressive Segmentation framework via next-scale mask prediction, termed AR-Seg, which progressively predicts the next-scale mask by explicitly modeling dependencies across all previous scales within a unified architecture. AR-Seg introduces three innovations: (1) a multi-scale mask autoencoder that quantizes the mask into multi-scale token maps to capture hierarchical anatomical structures, (2) a next-scale autoregressive mechanism that progressively predicts next-scale masks to enable sufficient inter-scale dependencies, and (3) a consensus-aggregation strategy that combines multiple sampled results to generate a more accurate mask, further improving segmentation robustness. Extensive experimental results on two benchmark datasets with different modalities demonstrate that AR-Seg outperforms state-of-the-art methods while explicitly visualizing the intermediate coarse-to-fine segmentation process.
no_new_dataset
0.949669
2502.20785
Hyewon Jeon
Hyewon Jeon, Jay-Yoon Lee
GraphCheck: Multi-Path Fact-Checking with Entity-Relationship Graphs
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
Automated fact-checking aims to assess the truthfulness of text based on relevant evidence, yet verifying complex claims requiring multi-hop reasoning remains a significant challenge. We propose GraphCheck, a novel framework that converts claims into entity-relationship graphs for comprehensive verification. By identifying relation between explicit entities and latent entities across multiple paths, GraphCheck enhances the adaptability and robustness of verification. Furthermore, we introduce DP-GraphCheck, a two-stage variant that improves performance by incorporating direct prompting as an initial filtering step. Experiments on the HOVER and EX-FEVER datasets show that our approach outperforms existing methods, particularly in multi-hop reasoning tasks. Furthermore, our two-stage framework generalizes well to other fact-checking pipelines, demonstrating its versatility.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 07:06:19 GMT" } ]
2025-03-03T00:00:00
[ [ "Jeon", "Hyewon", "" ], [ "Lee", "Jay-Yoon", "" ] ]
TITLE: GraphCheck: Multi-Path Fact-Checking with Entity-Relationship Graphs ABSTRACT: Automated fact-checking aims to assess the truthfulness of text based on relevant evidence, yet verifying complex claims requiring multi-hop reasoning remains a significant challenge. We propose GraphCheck, a novel framework that converts claims into entity-relationship graphs for comprehensive verification. By identifying relation between explicit entities and latent entities across multiple paths, GraphCheck enhances the adaptability and robustness of verification. Furthermore, we introduce DP-GraphCheck, a two-stage variant that improves performance by incorporating direct prompting as an initial filtering step. Experiments on the HOVER and EX-FEVER datasets show that our approach outperforms existing methods, particularly in multi-hop reasoning tasks. Furthermore, our two-stage framework generalizes well to other fact-checking pipelines, demonstrating its versatility.
no_new_dataset
0.944434
2502.20793
Amir Jahangiri
Amir Jahangiri, Tatiana Agback, Ulrika Brath and Vladislav Orekhov
Towards Ultimate NMR Resolution with Deep Learning
null
null
null
null
physics.bio-ph cs.LG physics.data-an
http://creativecommons.org/licenses/by-nc-nd/4.0/
In multidimensional NMR spectroscopy, practical resolution is defined as the ability to distinguish and accurately determine signal positions against a background of overlapping peaks, thermal noise, and spectral artifacts. In the pursuit of ultimate resolution, we introduce Peak Probability Presentations ($P^3$)- a statistical spectral representation that assigns a probability to each spectral point, indicating the likelihood of a peak maximum occurring at that location. The mapping between the spectrum and $P^3$ is achieved using MR-Ai, a physics-inspired deep learning neural network architecture, designed to handle multidimensional NMR spectra. Furthermore, we demonstrate that MR-Ai enables coprocessing of multiple spectra, facilitating direct information exchange between datasets. This feature significantly enhances spectral quality, particularly in cases of highly sparse sampling. Performance of MR-Ai and high value of the $P^3$ are demonstrated on the synthetic data and spectra of Tau, MATL1, Calmodulin, and several other proteins.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 07:20:25 GMT" } ]
2025-03-03T00:00:00
[ [ "Jahangiri", "Amir", "" ], [ "Agback", "Tatiana", "" ], [ "Brath", "Ulrika", "" ], [ "Orekhov", "Vladislav", "" ] ]
TITLE: Towards Ultimate NMR Resolution with Deep Learning ABSTRACT: In multidimensional NMR spectroscopy, practical resolution is defined as the ability to distinguish and accurately determine signal positions against a background of overlapping peaks, thermal noise, and spectral artifacts. In the pursuit of ultimate resolution, we introduce Peak Probability Presentations ($P^3$)- a statistical spectral representation that assigns a probability to each spectral point, indicating the likelihood of a peak maximum occurring at that location. The mapping between the spectrum and $P^3$ is achieved using MR-Ai, a physics-inspired deep learning neural network architecture, designed to handle multidimensional NMR spectra. Furthermore, we demonstrate that MR-Ai enables coprocessing of multiple spectra, facilitating direct information exchange between datasets. This feature significantly enhances spectral quality, particularly in cases of highly sparse sampling. Performance of MR-Ai and high value of the $P^3$ are demonstrated on the synthetic data and spectra of Tau, MATL1, Calmodulin, and several other proteins.
no_new_dataset
0.951233
2502.20803
Masoumeh Chapariniya
Masoumeh Chapariniya, Hossein Ranjbar, Teodora Vukovic, Sarah Ebling, Volker Dellwo
Two-Stream Spatial-Temporal Transformer Framework for Person Identification via Natural Conversational Keypoints
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
In the age of AI-driven generative technologies, traditional biometric recognition systems face unprecedented challenges, particularly from sophisticated deepfake and face reenactment techniques. In this study, we propose a Two-Stream Spatial-Temporal Transformer Framework for person identification using upper body keypoints visible during online conversations, which we term conversational keypoints. Our framework processes both spatial relationships between keypoints and their temporal evolution through two specialized branches: a Spatial Transformer (STR) that learns distinctive structural patterns in keypoint configurations, and a Temporal Transformer (TTR) that captures sequential motion patterns. Using the state-of-the-art Sapiens pose estimator, we extract 133 keypoints (based on COCO-WholeBody format) representing facial features, head pose, and hand positions. The framework was evaluated on a dataset of 114 individuals engaged in natural conversations, achieving recognition accuracies of 80.12% for the spatial stream, 63.61% for the temporal stream. We then explored two fusion strategies: a shared loss function approach achieving 82.22% accuracy, and a feature-level fusion method that concatenates feature maps from both streams, significantly improving performance to 94.86%. By jointly modeling both static anatomical relationships and dynamic movement patterns, our approach learns comprehensive identity signatures that are more robust to spoofing than traditional appearance-based methods.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 07:38:48 GMT" } ]
2025-03-03T00:00:00
[ [ "Chapariniya", "Masoumeh", "" ], [ "Ranjbar", "Hossein", "" ], [ "Vukovic", "Teodora", "" ], [ "Ebling", "Sarah", "" ], [ "Dellwo", "Volker", "" ] ]
TITLE: Two-Stream Spatial-Temporal Transformer Framework for Person Identification via Natural Conversational Keypoints ABSTRACT: In the age of AI-driven generative technologies, traditional biometric recognition systems face unprecedented challenges, particularly from sophisticated deepfake and face reenactment techniques. In this study, we propose a Two-Stream Spatial-Temporal Transformer Framework for person identification using upper body keypoints visible during online conversations, which we term conversational keypoints. Our framework processes both spatial relationships between keypoints and their temporal evolution through two specialized branches: a Spatial Transformer (STR) that learns distinctive structural patterns in keypoint configurations, and a Temporal Transformer (TTR) that captures sequential motion patterns. Using the state-of-the-art Sapiens pose estimator, we extract 133 keypoints (based on COCO-WholeBody format) representing facial features, head pose, and hand positions. The framework was evaluated on a dataset of 114 individuals engaged in natural conversations, achieving recognition accuracies of 80.12% for the spatial stream, 63.61% for the temporal stream. We then explored two fusion strategies: a shared loss function approach achieving 82.22% accuracy, and a feature-level fusion method that concatenates feature maps from both streams, significantly improving performance to 94.86%. By jointly modeling both static anatomical relationships and dynamic movement patterns, our approach learns comprehensive identity signatures that are more robust to spoofing than traditional appearance-based methods.
no_new_dataset
0.94868
2502.20806
Faisal Mohammad
Faisal Mohammad, Duksan Ryu
Multimodal Learning for Just-In-Time Software Defect Prediction in Autonomous Driving Systems
9
null
null
null
cs.SE cs.AI
http://creativecommons.org/licenses/by/4.0/
In recent years, the rise of autonomous driving technologies has highlighted the critical importance of reliable software for ensuring safety and performance. This paper proposes a novel approach for just-in-time software defect prediction (JIT-SDP) in autonomous driving software systems using multimodal learning. The proposed model leverages the multimodal transformers in which the pre-trained transformers and a combining module deal with the multiple data modalities of the software system datasets such as code features, change metrics, and contextual information. The key point for adapting multimodal learning is to utilize the attention mechanism between the different data modalities such as text, numerical, and categorical. In the combining module, the output of a transformer model on text data and tabular features containing categorical and numerical data are combined to produce the predictions using the fully connected layers. Experiments conducted on three open-source autonomous driving system software projects collected from the GitHub repository (Apollo, Carla, and Donkeycar) demonstrate that the proposed approach significantly outperforms state-of-the-art deep learning and machine learning models regarding evaluation metrics. Our findings highlight the potential of multimodal learning to enhance the reliability and safety of autonomous driving software through improved defect prediction.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 07:45:10 GMT" } ]
2025-03-03T00:00:00
[ [ "Mohammad", "Faisal", "" ], [ "Ryu", "Duksan", "" ] ]
TITLE: Multimodal Learning for Just-In-Time Software Defect Prediction in Autonomous Driving Systems ABSTRACT: In recent years, the rise of autonomous driving technologies has highlighted the critical importance of reliable software for ensuring safety and performance. This paper proposes a novel approach for just-in-time software defect prediction (JIT-SDP) in autonomous driving software systems using multimodal learning. The proposed model leverages the multimodal transformers in which the pre-trained transformers and a combining module deal with the multiple data modalities of the software system datasets such as code features, change metrics, and contextual information. The key point for adapting multimodal learning is to utilize the attention mechanism between the different data modalities such as text, numerical, and categorical. In the combining module, the output of a transformer model on text data and tabular features containing categorical and numerical data are combined to produce the predictions using the fully connected layers. Experiments conducted on three open-source autonomous driving system software projects collected from the GitHub repository (Apollo, Carla, and Donkeycar) demonstrate that the proposed approach significantly outperforms state-of-the-art deep learning and machine learning models regarding evaluation metrics. Our findings highlight the potential of multimodal learning to enhance the reliability and safety of autonomous driving software through improved defect prediction.
no_new_dataset
0.949106
2502.20811
Xiao Wang
Xiao Wang, Jingyun Hua, Weihong Lin, Yuanxing Zhang, Fuzheng Zhang, Jianlong Wu, Di Zhang, Liqiang Nie
HAIC: Improving Human Action Understanding and Generation with Better Captions for Multi-modal Large Language Models
null
null
null
null
cs.CV cs.CL cs.MM
http://creativecommons.org/licenses/by/4.0/
Recent Multi-modal Large Language Models (MLLMs) have made great progress in video understanding. However, their performance on videos involving human actions is still limited by the lack of high-quality data. To address this, we introduce a two-stage data annotation pipeline. First, we design strategies to accumulate videos featuring clear human actions from the Internet. Second, videos are annotated in a standardized caption format that uses human attributes to distinguish individuals and chronologically details their actions and interactions. Through this pipeline, we curate two datasets, namely HAICTrain and HAICBench. \textbf{HAICTrain} comprises 126K video-caption pairs generated by Gemini-Pro and verified for training purposes. Meanwhile, \textbf{HAICBench} includes 500 manually annotated video-caption pairs and 1,400 QA pairs, for a comprehensive evaluation of human action understanding. Experimental results demonstrate that training with HAICTrain not only significantly enhances human understanding abilities across 4 benchmarks, but can also improve text-to-video generation results. Both the HAICTrain and HAICBench are released at https://huggingface.co/datasets/KuaishouHAIC/HAIC.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 07:53:40 GMT" } ]
2025-03-03T00:00:00
[ [ "Wang", "Xiao", "" ], [ "Hua", "Jingyun", "" ], [ "Lin", "Weihong", "" ], [ "Zhang", "Yuanxing", "" ], [ "Zhang", "Fuzheng", "" ], [ "Wu", "Jianlong", "" ], [ "Zhang", "Di", "" ], [ "Nie", "Liqiang", "" ] ]
TITLE: HAIC: Improving Human Action Understanding and Generation with Better Captions for Multi-modal Large Language Models ABSTRACT: Recent Multi-modal Large Language Models (MLLMs) have made great progress in video understanding. However, their performance on videos involving human actions is still limited by the lack of high-quality data. To address this, we introduce a two-stage data annotation pipeline. First, we design strategies to accumulate videos featuring clear human actions from the Internet. Second, videos are annotated in a standardized caption format that uses human attributes to distinguish individuals and chronologically details their actions and interactions. Through this pipeline, we curate two datasets, namely HAICTrain and HAICBench. \textbf{HAICTrain} comprises 126K video-caption pairs generated by Gemini-Pro and verified for training purposes. Meanwhile, \textbf{HAICBench} includes 500 manually annotated video-caption pairs and 1,400 QA pairs, for a comprehensive evaluation of human action understanding. Experimental results demonstrate that training with HAICTrain not only significantly enhances human understanding abilities across 4 benchmarks, but can also improve text-to-video generation results. Both the HAICTrain and HAICBench are released at https://huggingface.co/datasets/KuaishouHAIC/HAIC.
new_dataset
0.95846
2502.20824
Fadeel Sher Khan
Fadeel Sher Khan, Joshua Ebenezer, Hamid Sheikh, Seok-Jun Lee
MFSR-GAN: Multi-Frame Super-Resolution with Handheld Motion Modeling
8 pages, 6 figures
null
null
null
cs.CV eess.IV
http://creativecommons.org/licenses/by-nc-nd/4.0/
Smartphone cameras have become ubiquitous imaging tools, yet their small sensors and compact optics often limit spatial resolution and introduce distortions. Combining information from multiple low-resolution (LR) frames to produce a high-resolution (HR) image has been explored to overcome the inherent limitations of smartphone cameras. Despite the promise of multi-frame super-resolution (MFSR), current approaches are hindered by datasets that fail to capture the characteristic noise and motion patterns found in real-world handheld burst images. In this work, we address this gap by introducing a novel synthetic data engine that uses multi-exposure static images to synthesize LR-HR training pairs while preserving sensor-specific noise characteristics and image motion found during handheld burst photography. We also propose MFSR-GAN: a multi-scale RAW-to-RGB network for MFSR. Compared to prior approaches, MFSR-GAN emphasizes a "base frame" throughout its architecture to mitigate artifacts. Experimental results on both synthetic and real data demonstrates that MFSR-GAN trained with our synthetic engine yields sharper, more realistic reconstructions than existing methods for real-world MFSR.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 08:11:03 GMT" } ]
2025-03-03T00:00:00
[ [ "Khan", "Fadeel Sher", "" ], [ "Ebenezer", "Joshua", "" ], [ "Sheikh", "Hamid", "" ], [ "Lee", "Seok-Jun", "" ] ]
TITLE: MFSR-GAN: Multi-Frame Super-Resolution with Handheld Motion Modeling ABSTRACT: Smartphone cameras have become ubiquitous imaging tools, yet their small sensors and compact optics often limit spatial resolution and introduce distortions. Combining information from multiple low-resolution (LR) frames to produce a high-resolution (HR) image has been explored to overcome the inherent limitations of smartphone cameras. Despite the promise of multi-frame super-resolution (MFSR), current approaches are hindered by datasets that fail to capture the characteristic noise and motion patterns found in real-world handheld burst images. In this work, we address this gap by introducing a novel synthetic data engine that uses multi-exposure static images to synthesize LR-HR training pairs while preserving sensor-specific noise characteristics and image motion found during handheld burst photography. We also propose MFSR-GAN: a multi-scale RAW-to-RGB network for MFSR. Compared to prior approaches, MFSR-GAN emphasizes a "base frame" throughout its architecture to mitigate artifacts. Experimental results on both synthetic and real data demonstrates that MFSR-GAN trained with our synthetic engine yields sharper, more realistic reconstructions than existing methods for real-world MFSR.
no_new_dataset
0.952397
2502.20850
Anh Tien Nguyen
Anh Tien Nguyen, Keunho Byeon, Kyungeun Kim, Jin Tae Kwak
VLEER: Vision and Language Embeddings for Explainable Whole Slide Image Representation
Under review
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
Recent advances in vision-language models (VLMs) have shown remarkable potential in bridging visual and textual modalities. In computational pathology, domain-specific VLMs, which are pre-trained on extensive histopathology image-text datasets, have succeeded in various downstream tasks. However, existing research has primarily focused on the pre-training process and direct applications of VLMs on the patch level, leaving their great potential for whole slide image (WSI) applications unexplored. In this study, we hypothesize that pre-trained VLMs inherently capture informative and interpretable WSI representations through quantitative feature extraction. To validate this hypothesis, we introduce Vision and Language Embeddings for Explainable WSI Representation (VLEER), a novel method designed to leverage VLMs for WSI representation. We systematically evaluate VLEER on three pathological WSI datasets, proving its better performance in WSI analysis compared to conventional vision features. More importantly, VLEER offers the unique advantage of interpretability, enabling direct human-readable insights into the results by leveraging the textual modality for detailed pathology annotations, providing clear reasoning for WSI-level pathology downstream tasks.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 08:49:03 GMT" } ]
2025-03-03T00:00:00
[ [ "Nguyen", "Anh Tien", "" ], [ "Byeon", "Keunho", "" ], [ "Kim", "Kyungeun", "" ], [ "Kwak", "Jin Tae", "" ] ]
TITLE: VLEER: Vision and Language Embeddings for Explainable Whole Slide Image Representation ABSTRACT: Recent advances in vision-language models (VLMs) have shown remarkable potential in bridging visual and textual modalities. In computational pathology, domain-specific VLMs, which are pre-trained on extensive histopathology image-text datasets, have succeeded in various downstream tasks. However, existing research has primarily focused on the pre-training process and direct applications of VLMs on the patch level, leaving their great potential for whole slide image (WSI) applications unexplored. In this study, we hypothesize that pre-trained VLMs inherently capture informative and interpretable WSI representations through quantitative feature extraction. To validate this hypothesis, we introduce Vision and Language Embeddings for Explainable WSI Representation (VLEER), a novel method designed to leverage VLMs for WSI representation. We systematically evaluate VLEER on three pathological WSI datasets, proving its better performance in WSI analysis compared to conventional vision features. More importantly, VLEER offers the unique advantage of interpretability, enabling direct human-readable insights into the results by leveraging the textual modality for detailed pathology annotations, providing clear reasoning for WSI-level pathology downstream tasks.
no_new_dataset
0.944125
2502.20852
Rongchang Lu
Rongchang Lu, Bingcheng Liao, Haowen Hou, Jiahang Lv and Xin Hai
Delta-WKV: A Novel Meta-in-Context Learner for MRI Super-Resolution
This paper has been published to MICCAI 2025. Feel free to contact on [email protected]
null
null
null
eess.IV cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Magnetic Resonance Imaging (MRI) Super-Resolution (SR) addresses the challenges such as long scan times and expensive equipment by enhancing image resolution from low-quality inputs acquired in shorter scan times in clinical settings. However, current SR techniques still have problems such as limited ability to capture both local and global static patterns effectively and efficiently. To address these limitations, we propose Delta-WKV, a novel MRI super-resolution model that combines Meta-in-Context Learning (MiCL) with the Delta rule to better recognize both local and global patterns in MRI images. This approach allows Delta-WKV to adjust weights dynamically during inference, improving pattern recognition with fewer parameters and less computational effort, without using state-space modeling. Additionally, inspired by Receptance Weighted Key Value (RWKV), Delta-WKV uses a quad-directional scanning mechanism with time-mixing and channel-mixing structures to capture long-range dependencies while maintaining high-frequency details. Tests on the IXI and fastMRI datasets show that Delta-WKV outperforms existing methods, improving PSNR by 0.06 dB and SSIM by 0.001, while reducing training and inference times by over 15\%. These results demonstrate its efficiency and potential for clinical use with large datasets and high-resolution imaging.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 08:49:46 GMT" } ]
2025-03-03T00:00:00
[ [ "Lu", "Rongchang", "" ], [ "Liao", "Bingcheng", "" ], [ "Hou", "Haowen", "" ], [ "Lv", "Jiahang", "" ], [ "Hai", "Xin", "" ] ]
TITLE: Delta-WKV: A Novel Meta-in-Context Learner for MRI Super-Resolution ABSTRACT: Magnetic Resonance Imaging (MRI) Super-Resolution (SR) addresses the challenges such as long scan times and expensive equipment by enhancing image resolution from low-quality inputs acquired in shorter scan times in clinical settings. However, current SR techniques still have problems such as limited ability to capture both local and global static patterns effectively and efficiently. To address these limitations, we propose Delta-WKV, a novel MRI super-resolution model that combines Meta-in-Context Learning (MiCL) with the Delta rule to better recognize both local and global patterns in MRI images. This approach allows Delta-WKV to adjust weights dynamically during inference, improving pattern recognition with fewer parameters and less computational effort, without using state-space modeling. Additionally, inspired by Receptance Weighted Key Value (RWKV), Delta-WKV uses a quad-directional scanning mechanism with time-mixing and channel-mixing structures to capture long-range dependencies while maintaining high-frequency details. Tests on the IXI and fastMRI datasets show that Delta-WKV outperforms existing methods, improving PSNR by 0.06 dB and SSIM by 0.001, while reducing training and inference times by over 15\%. These results demonstrate its efficiency and potential for clinical use with large datasets and high-resolution imaging.
no_new_dataset
0.950457
2502.20855
Jonathan Drechsel
Jonathan Drechsel, Anja Reusch, Steffen Herbold
MAMUT: A Novel Framework for Modifying Mathematical Formulas for the Generation of Specialized Datasets for Language Model Training
null
null
null
null
cs.CL cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Mathematical formulas are a fundamental and widely used component in various scientific fields, serving as a universal language for expressing complex concepts and relationships. While state-of-the-art transformer models excel in processing and understanding natural language, they encounter challenges with mathematical notation, which involves a complex structure and diverse representations. This study focuses on the development of specialized training datasets to enhance the encoding of mathematical content. We introduce Math Mutator (MAMUT), a framework capable of generating equivalent and falsified versions of a given mathematical formula in LaTeX notation, effectively capturing the mathematical variety in notation of the same concept. Based on MAMUT, we have generated four large mathematical datasets containing diverse notation, which can be used to train language models with enhanced mathematical embeddings.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 08:53:42 GMT" } ]
2025-03-03T00:00:00
[ [ "Drechsel", "Jonathan", "" ], [ "Reusch", "Anja", "" ], [ "Herbold", "Steffen", "" ] ]
TITLE: MAMUT: A Novel Framework for Modifying Mathematical Formulas for the Generation of Specialized Datasets for Language Model Training ABSTRACT: Mathematical formulas are a fundamental and widely used component in various scientific fields, serving as a universal language for expressing complex concepts and relationships. While state-of-the-art transformer models excel in processing and understanding natural language, they encounter challenges with mathematical notation, which involves a complex structure and diverse representations. This study focuses on the development of specialized training datasets to enhance the encoding of mathematical content. We introduce Math Mutator (MAMUT), a framework capable of generating equivalent and falsified versions of a given mathematical formula in LaTeX notation, effectively capturing the mathematical variety in notation of the same concept. Based on MAMUT, we have generated four large mathematical datasets containing diverse notation, which can be used to train language models with enhanced mathematical embeddings.
new_dataset
0.949529
2502.20857
Hyeonuk Nam
Hyeonuk Nam, Yong-Hwa Park
JiTTER: Jigsaw Temporal Transformer for Event Reconstruction for Self-Supervised Sound Event Detection
null
null
null
null
eess.AS cs.SD
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sound event detection (SED) has significantly benefited from self-supervised learning (SSL) approaches, particularly masked audio transformer for SED (MAT-SED), which leverages masked block prediction to reconstruct missing audio segments. However, while effective in capturing global dependencies, masked block prediction disrupts transient sound events and lacks explicit enforcement of temporal order, making it less suitable for fine-grained event boundary detection. To address these limitations, we propose JiTTER (Jigsaw Temporal Transformer for Event Reconstruction), an SSL framework designed to enhance temporal modeling in transformer-based SED. JiTTER introduces a hierarchical temporal shuffle reconstruction strategy, where audio sequences are randomly shuffled at both the block-level and frame-level, forcing the model to reconstruct the correct temporal order. This pretraining objective encourages the model to learn both global event structures and fine-grained transient details, improving its ability to detect events with sharp onset-offset characteristics. Additionally, we incorporate noise injection during block shuffle, providing a subtle perturbation mechanism that further regularizes feature learning and enhances model robustness. Experimental results on the DESED dataset demonstrate that JiTTER outperforms MAT-SED, achieving a 5.89% improvement in PSDS, highlighting the effectiveness of explicit temporal reasoning in SSL-based SED. Our findings suggest that structured temporal reconstruction tasks, rather than simple masked prediction, offer a more effective pretraining paradigm for sound event representation learning.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 08:55:20 GMT" } ]
2025-03-03T00:00:00
[ [ "Nam", "Hyeonuk", "" ], [ "Park", "Yong-Hwa", "" ] ]
TITLE: JiTTER: Jigsaw Temporal Transformer for Event Reconstruction for Self-Supervised Sound Event Detection ABSTRACT: Sound event detection (SED) has significantly benefited from self-supervised learning (SSL) approaches, particularly masked audio transformer for SED (MAT-SED), which leverages masked block prediction to reconstruct missing audio segments. However, while effective in capturing global dependencies, masked block prediction disrupts transient sound events and lacks explicit enforcement of temporal order, making it less suitable for fine-grained event boundary detection. To address these limitations, we propose JiTTER (Jigsaw Temporal Transformer for Event Reconstruction), an SSL framework designed to enhance temporal modeling in transformer-based SED. JiTTER introduces a hierarchical temporal shuffle reconstruction strategy, where audio sequences are randomly shuffled at both the block-level and frame-level, forcing the model to reconstruct the correct temporal order. This pretraining objective encourages the model to learn both global event structures and fine-grained transient details, improving its ability to detect events with sharp onset-offset characteristics. Additionally, we incorporate noise injection during block shuffle, providing a subtle perturbation mechanism that further regularizes feature learning and enhances model robustness. Experimental results on the DESED dataset demonstrate that JiTTER outperforms MAT-SED, achieving a 5.89% improvement in PSDS, highlighting the effectiveness of explicit temporal reasoning in SSL-based SED. Our findings suggest that structured temporal reconstruction tasks, rather than simple masked prediction, offer a more effective pretraining paradigm for sound event representation learning.
no_new_dataset
0.953144
2502.20858
Xiaochuan Liu
Xiaochuan Liu, Xin Cheng, Yuchong Sun, Xiaoxue Wu, Ruihua Song, Hao Sun, Denghao Zhang
EyEar: Learning Audio Synchronized Human Gaze Trajectory Based on Physics-Informed Dynamics
null
null
null
null
cs.MM
http://creativecommons.org/licenses/by/4.0/
Imitating how humans move their gaze in a visual scene is a vital research problem for both visual understanding and psychology, kindling crucial applications such as building alive virtual characters. Previous studies aim to predict gaze trajectories when humans are free-viewing an image, searching for required targets, or looking for clues to answer questions in an image. While these tasks focus on visual-centric scenarios, humans move their gaze also along with audio signal inputs in more common scenarios. To fill this gap, we introduce a new task that predicts human gaze trajectories in a visual scene with synchronized audio inputs and provide a new dataset containing 20k gaze points from 8 subjects. To effectively integrate audio information and simulate the dynamic process of human gaze motion, we propose a novel learning framework called EyEar (Eye moving while Ear listening) based on physics-informed dynamics, which considers three key factors to predict gazes: eye inherent motion tendency, vision salient attraction, and audio semantic attraction. We also propose a probability density score to overcome the high individual variability of gaze trajectories, thereby improving the stabilization of optimization and the reliability of the evaluation. Experimental results show that EyEar outperforms all the baselines in the context of all evaluation metrics, thanks to the proposed components in the learning model.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:01:30 GMT" } ]
2025-03-03T00:00:00
[ [ "Liu", "Xiaochuan", "" ], [ "Cheng", "Xin", "" ], [ "Sun", "Yuchong", "" ], [ "Wu", "Xiaoxue", "" ], [ "Song", "Ruihua", "" ], [ "Sun", "Hao", "" ], [ "Zhang", "Denghao", "" ] ]
TITLE: EyEar: Learning Audio Synchronized Human Gaze Trajectory Based on Physics-Informed Dynamics ABSTRACT: Imitating how humans move their gaze in a visual scene is a vital research problem for both visual understanding and psychology, kindling crucial applications such as building alive virtual characters. Previous studies aim to predict gaze trajectories when humans are free-viewing an image, searching for required targets, or looking for clues to answer questions in an image. While these tasks focus on visual-centric scenarios, humans move their gaze also along with audio signal inputs in more common scenarios. To fill this gap, we introduce a new task that predicts human gaze trajectories in a visual scene with synchronized audio inputs and provide a new dataset containing 20k gaze points from 8 subjects. To effectively integrate audio information and simulate the dynamic process of human gaze motion, we propose a novel learning framework called EyEar (Eye moving while Ear listening) based on physics-informed dynamics, which considers three key factors to predict gazes: eye inherent motion tendency, vision salient attraction, and audio semantic attraction. We also propose a probability density score to overcome the high individual variability of gaze trajectories, thereby improving the stabilization of optimization and the reliability of the evaluation. Experimental results show that EyEar outperforms all the baselines in the context of all evaluation metrics, thanks to the proposed components in the learning model.
new_dataset
0.955527
2502.20862
Ho Fai Po
Ho Fai Po, Akke Mats Houben, Anna-Christina Haeb, Yordan P. Raykov, Daniel Tornero, Jordi Soriano, David Saad
Analysis of Evolving Cortical Neuronal Networks Using Visual Informatics
null
null
null
null
q-bio.NC cond-mat.dis-nn physics.data-an physics.soc-ph
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Understanding the nature of the changes exhibited by evolving neuronal dynamics from high-dimensional activity data is essential for advancing neuroscience, particularly in the study of neuronal network development and the pathophysiology of neurological disorders. This work examines how advanced dimensionality reduction techniques can efficiently summarize and enhance our understanding of the development of neuronal networks over time and in response to stimulation. We develop a framework based on the Minimum-Distortion Embedding (MDE) methods and demonstrate how MDE outperforms better known benchmarks based on Principal Component Analysis (PCA) and t-distributed Stochastic Neighbor Embedding (t-SNE) by effectively preserving both global structures and local relationships within complex neuronal datasets. Our \emph{in silico} experiments reveal MDE's capability to capture the evolving connectivity patterns of simulated neuronal networks, illustrating a clear trajectory tracking the simulated network development. Complementary \emph{in vitro} experiments further validate MDE's advantages, highlighting its ability to identify behavioral differences and connectivity changes in neuronal cultures over a 35-day observation period. Additionally, we explore the effects of stimulation on neuronal activity, providing valuable insights into the plasticity and learning mechanisms of neuronal networks. Our findings underscore the importance of metric selection in dimensionality reduction, showing that correlation metrics yield more meaningful embeddings compared to Euclidean distance. The implications of this research extend to various areas, including the potential development of therapeutic intervention strategies for neurological disorders, and the identification of distinct phases of neuronal activity for advancing cortical-based computing devices.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:02:23 GMT" } ]
2025-03-03T00:00:00
[ [ "Po", "Ho Fai", "" ], [ "Houben", "Akke Mats", "" ], [ "Haeb", "Anna-Christina", "" ], [ "Raykov", "Yordan P.", "" ], [ "Tornero", "Daniel", "" ], [ "Soriano", "Jordi", "" ], [ "Saad", "David", "" ] ]
TITLE: Analysis of Evolving Cortical Neuronal Networks Using Visual Informatics ABSTRACT: Understanding the nature of the changes exhibited by evolving neuronal dynamics from high-dimensional activity data is essential for advancing neuroscience, particularly in the study of neuronal network development and the pathophysiology of neurological disorders. This work examines how advanced dimensionality reduction techniques can efficiently summarize and enhance our understanding of the development of neuronal networks over time and in response to stimulation. We develop a framework based on the Minimum-Distortion Embedding (MDE) methods and demonstrate how MDE outperforms better known benchmarks based on Principal Component Analysis (PCA) and t-distributed Stochastic Neighbor Embedding (t-SNE) by effectively preserving both global structures and local relationships within complex neuronal datasets. Our \emph{in silico} experiments reveal MDE's capability to capture the evolving connectivity patterns of simulated neuronal networks, illustrating a clear trajectory tracking the simulated network development. Complementary \emph{in vitro} experiments further validate MDE's advantages, highlighting its ability to identify behavioral differences and connectivity changes in neuronal cultures over a 35-day observation period. Additionally, we explore the effects of stimulation on neuronal activity, providing valuable insights into the plasticity and learning mechanisms of neuronal networks. Our findings underscore the importance of metric selection in dimensionality reduction, showing that correlation metrics yield more meaningful embeddings compared to Euclidean distance. The implications of this research extend to various areas, including the potential development of therapeutic intervention strategies for neurological disorders, and the identification of distinct phases of neuronal activity for advancing cortical-based computing devices.
no_new_dataset
0.949342
2502.20864
Mohammad Farhansyah Rifqi
Mohammad Rifqi Farhansyah, Iwan Darmawan, Adryan Kusumawardhana, Genta Indra Winata, Alham Fikri Aji, Derry Tanti Wijaya
Do Language Models Understand Honorific Systems in Javanese?
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by-nc-sa/4.0/
The Javanese language features a complex system of honorifics that vary according to the social status of the speaker, listener, and referent. Despite its cultural and linguistic significance, there has been limited progress in developing a comprehensive corpus to capture these variations for natural language processing (NLP) tasks. In this paper, we present Unggah-Ungguh, a carefully curated dataset designed to encapsulate the nuances of Unggah-Ungguh Basa, the Javanese speech etiquette framework that dictates the choice of words and phrases based on social hierarchy and context. Using Unggah-Ungguh, we assess the ability of language models (LMs) to process various levels of Javanese honorifics through classification and machine translation tasks. To further evaluate cross-lingual LMs, we conduct machine translation experiments between Javanese (at specific honorific levels) and Indonesian. Additionally, we explore whether LMs can generate contextually appropriate Javanese honorifics in conversation tasks, where the honorific usage should align with the social role and contextual cues. Our findings indicate that current LMs struggle with most honorific levels, exhibitinga bias toward certain honorific tiers.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:05:35 GMT" } ]
2025-03-03T00:00:00
[ [ "Farhansyah", "Mohammad Rifqi", "" ], [ "Darmawan", "Iwan", "" ], [ "Kusumawardhana", "Adryan", "" ], [ "Winata", "Genta Indra", "" ], [ "Aji", "Alham Fikri", "" ], [ "Wijaya", "Derry Tanti", "" ] ]
TITLE: Do Language Models Understand Honorific Systems in Javanese? ABSTRACT: The Javanese language features a complex system of honorifics that vary according to the social status of the speaker, listener, and referent. Despite its cultural and linguistic significance, there has been limited progress in developing a comprehensive corpus to capture these variations for natural language processing (NLP) tasks. In this paper, we present Unggah-Ungguh, a carefully curated dataset designed to encapsulate the nuances of Unggah-Ungguh Basa, the Javanese speech etiquette framework that dictates the choice of words and phrases based on social hierarchy and context. Using Unggah-Ungguh, we assess the ability of language models (LMs) to process various levels of Javanese honorifics through classification and machine translation tasks. To further evaluate cross-lingual LMs, we conduct machine translation experiments between Javanese (at specific honorific levels) and Indonesian. Additionally, we explore whether LMs can generate contextually appropriate Javanese honorifics in conversation tasks, where the honorific usage should align with the social role and contextual cues. Our findings indicate that current LMs struggle with most honorific levels, exhibitinga bias toward certain honorific tiers.
new_dataset
0.95846
2502.20869
Chunlin Zhong
Chunlin Zhong, Shuang Hao, Junhua Wu, Xiaona Chang, Jiwei Jiang, Xiu Nie, He Tang, Xiang Bai
PathVG: A New Benchmark and Dataset for Pathology Visual Grounding
10pages, 4figures
null
null
null
cs.CV
http://creativecommons.org/licenses/by-nc-nd/4.0/
With the rapid development of computational pathology, many AI-assisted diagnostic tasks have emerged. Cellular nuclei segmentation can segment various types of cells for downstream analysis, but it relies on predefined categories and lacks flexibility. Moreover, pathology visual question answering can perform image-level understanding but lacks region-level detection capability. To address this, we propose a new benchmark called Pathology Visual Grounding (PathVG), which aims to detect regions based on expressions with different attributes. To evaluate PathVG, we create a new dataset named RefPath which contains 27,610 images with 33,500 language-grounded boxes. Compared to visual grounding in other domains, PathVG presents pathological images at multi-scale and contains expressions with pathological knowledge. In the experimental study, we found that the biggest challenge was the implicit information underlying the pathological expressions. Based on this, we proposed Pathology Knowledge-enhanced Network (PKNet) as the baseline model for PathVG. PKNet leverages the knowledge-enhancement capabilities of Large Language Models (LLMs) to convert pathological terms with implicit information into explicit visual features, and fuses knowledge features with expression features through the designed Knowledge Fusion Module (KFM). The proposed method achieves state-of-the-art performance on the PathVG benchmark.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:13:01 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhong", "Chunlin", "" ], [ "Hao", "Shuang", "" ], [ "Wu", "Junhua", "" ], [ "Chang", "Xiaona", "" ], [ "Jiang", "Jiwei", "" ], [ "Nie", "Xiu", "" ], [ "Tang", "He", "" ], [ "Bai", "Xiang", "" ] ]
TITLE: PathVG: A New Benchmark and Dataset for Pathology Visual Grounding ABSTRACT: With the rapid development of computational pathology, many AI-assisted diagnostic tasks have emerged. Cellular nuclei segmentation can segment various types of cells for downstream analysis, but it relies on predefined categories and lacks flexibility. Moreover, pathology visual question answering can perform image-level understanding but lacks region-level detection capability. To address this, we propose a new benchmark called Pathology Visual Grounding (PathVG), which aims to detect regions based on expressions with different attributes. To evaluate PathVG, we create a new dataset named RefPath which contains 27,610 images with 33,500 language-grounded boxes. Compared to visual grounding in other domains, PathVG presents pathological images at multi-scale and contains expressions with pathological knowledge. In the experimental study, we found that the biggest challenge was the implicit information underlying the pathological expressions. Based on this, we proposed Pathology Knowledge-enhanced Network (PKNet) as the baseline model for PathVG. PKNet leverages the knowledge-enhancement capabilities of Large Language Models (LLMs) to convert pathological terms with implicit information into explicit visual features, and fuses knowledge features with expression features through the designed Knowledge Fusion Module (KFM). The proposed method achieves state-of-the-art performance on the PathVG benchmark.
new_dataset
0.95594
2502.20877
Haozhong Sun
Haozhong Sun, Zhongsen Li, Chenlin Du, Haokun Li, Yajie Wang, Huijun Chen
Guiding Quantitative MRI Reconstruction with Phase-wise Uncertainty
Submitted to MICCAI2025
null
null
null
eess.IV cs.CV
http://creativecommons.org/licenses/by-nc-nd/4.0/
Quantitative magnetic resonance imaging (qMRI) requires multi-phase acqui-sition, often relying on reduced data sampling and reconstruction algorithms to accelerate scans, which inherently poses an ill-posed inverse problem. While many studies focus on measuring uncertainty during this process, few explore how to leverage it to enhance reconstruction performance. In this paper, we in-troduce PUQ, a novel approach that pioneers the use of uncertainty infor-mation for qMRI reconstruction. PUQ employs a two-stage reconstruction and parameter fitting framework, where phase-wise uncertainty is estimated during reconstruction and utilized in the fitting stage. This design allows uncertainty to reflect the reliability of different phases and guide information integration during parameter fitting. We evaluated PUQ on in vivo T1 and T2 mapping datasets from healthy subjects. Compared to existing qMRI reconstruction methods, PUQ achieved the state-of-the-art performance in parameter map-pings, demonstrating the effectiveness of uncertainty guidance. Our code is available at https://anonymous.4open.science/r/PUQ-75B2/.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:21:01 GMT" } ]
2025-03-03T00:00:00
[ [ "Sun", "Haozhong", "" ], [ "Li", "Zhongsen", "" ], [ "Du", "Chenlin", "" ], [ "Li", "Haokun", "" ], [ "Wang", "Yajie", "" ], [ "Chen", "Huijun", "" ] ]
TITLE: Guiding Quantitative MRI Reconstruction with Phase-wise Uncertainty ABSTRACT: Quantitative magnetic resonance imaging (qMRI) requires multi-phase acqui-sition, often relying on reduced data sampling and reconstruction algorithms to accelerate scans, which inherently poses an ill-posed inverse problem. While many studies focus on measuring uncertainty during this process, few explore how to leverage it to enhance reconstruction performance. In this paper, we in-troduce PUQ, a novel approach that pioneers the use of uncertainty infor-mation for qMRI reconstruction. PUQ employs a two-stage reconstruction and parameter fitting framework, where phase-wise uncertainty is estimated during reconstruction and utilized in the fitting stage. This design allows uncertainty to reflect the reliability of different phases and guide information integration during parameter fitting. We evaluated PUQ on in vivo T1 and T2 mapping datasets from healthy subjects. Compared to existing qMRI reconstruction methods, PUQ achieved the state-of-the-art performance in parameter map-pings, demonstrating the effectiveness of uncertainty guidance. Our code is available at https://anonymous.4open.science/r/PUQ-75B2/.
no_new_dataset
0.948917
2502.20879
Bj\"orn Braun
Bj\"orn Braun, Rayan Armani, Manuel Meier, Max Moebus, Christian Holz
egoPPG: Heart Rate Estimation from Eye-Tracking Cameras in Egocentric Systems to Benefit Downstream Vision Tasks
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by-nc-nd/4.0/
Egocentric vision systems aim to understand the spatial surroundings and the wearer's behavior inside it, including motions, activities, and interaction with objects. Since a person's attention and situational responses are influenced by their physiological state, egocentric systems must also detect this state for better context awareness. In this paper, we propose egoPPG, a novel task for egocentric vision systems to extract a person's heart rate (HR) as a key indicator of the wearer's physiological state from the system's built-in sensors (e.g., eye tracking videos). We then propose EgoPulseFormer, a method that solely takes eye-tracking video as input to estimate a person's photoplethysmogram (PPG) from areas around the eyes to track HR values-without requiring additional or dedicated hardware. We demonstrate the downstream benefit of EgoPulseFormer on EgoExo4D, where we find that augmenting existing models with tracked HR values improves proficiency estimation by 14%. To train and validate EgoPulseFormer, we collected a dataset of 13+ hours of eye-tracking videos from Project Aria and contact-based blood volume pulse signals as well as an electrocardiogram (ECG) for ground-truth HR values. 25 participants performed diverse everyday activities such as office work, cooking, dancing, and exercising, which induced significant natural motion and HR variation (44-164 bpm). Our model robustly estimates HR (MAE=8.82 bpm) and captures patterns (r=0.81). Our results show how egocentric systems may unify environmental and physiological tracking to better understand user actions and internal states.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:23:40 GMT" } ]
2025-03-03T00:00:00
[ [ "Braun", "Björn", "" ], [ "Armani", "Rayan", "" ], [ "Meier", "Manuel", "" ], [ "Moebus", "Max", "" ], [ "Holz", "Christian", "" ] ]
TITLE: egoPPG: Heart Rate Estimation from Eye-Tracking Cameras in Egocentric Systems to Benefit Downstream Vision Tasks ABSTRACT: Egocentric vision systems aim to understand the spatial surroundings and the wearer's behavior inside it, including motions, activities, and interaction with objects. Since a person's attention and situational responses are influenced by their physiological state, egocentric systems must also detect this state for better context awareness. In this paper, we propose egoPPG, a novel task for egocentric vision systems to extract a person's heart rate (HR) as a key indicator of the wearer's physiological state from the system's built-in sensors (e.g., eye tracking videos). We then propose EgoPulseFormer, a method that solely takes eye-tracking video as input to estimate a person's photoplethysmogram (PPG) from areas around the eyes to track HR values-without requiring additional or dedicated hardware. We demonstrate the downstream benefit of EgoPulseFormer on EgoExo4D, where we find that augmenting existing models with tracked HR values improves proficiency estimation by 14%. To train and validate EgoPulseFormer, we collected a dataset of 13+ hours of eye-tracking videos from Project Aria and contact-based blood volume pulse signals as well as an electrocardiogram (ECG) for ground-truth HR values. 25 participants performed diverse everyday activities such as office work, cooking, dancing, and exercising, which induced significant natural motion and HR variation (44-164 bpm). Our model robustly estimates HR (MAE=8.82 bpm) and captures patterns (r=0.81). Our results show how egocentric systems may unify environmental and physiological tracking to better understand user actions and internal states.
new_dataset
0.96128
2502.20885
Jhony Heriberto Giraldo Zuluaga
Amadou S. Sangare, Nicolas Dunou, Jhony H. Giraldo, Fragkiskos D. Malliaros
A Fused Gromov-Wasserstein Approach to Subgraph Contrastive Learning
null
Transactions on Machine Learning Research, 2025
null
null
cs.LG cs.AI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Self-supervised learning has become a key method for training deep learning models when labeled data is scarce or unavailable. While graph machine learning holds great promise across various domains, the design of effective pretext tasks for self-supervised graph representation learning remains challenging. Contrastive learning, a popular approach in graph self-supervised learning, leverages positive and negative pairs to compute a contrastive loss function. However, current graph contrastive learning methods often struggle to fully use structural patterns and node similarities. To address these issues, we present a new method called Fused Gromov Wasserstein Subgraph Contrastive Learning (FOSSIL). Our model integrates node-level and subgraph-level contrastive learning, seamlessly combining a standard node-level contrastive loss with the Fused Gromov-Wasserstein distance. This combination helps our method capture both node features and graph structure together. Importantly, our approach works well with both homophilic and heterophilic graphs and can dynamically create views for generating positive and negative pairs. Through extensive experiments on benchmark graph datasets, we show that FOSSIL outperforms or achieves competitive performance compared to current state-of-the-art methods.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:32:07 GMT" } ]
2025-03-03T00:00:00
[ [ "Sangare", "Amadou S.", "" ], [ "Dunou", "Nicolas", "" ], [ "Giraldo", "Jhony H.", "" ], [ "Malliaros", "Fragkiskos D.", "" ] ]
TITLE: A Fused Gromov-Wasserstein Approach to Subgraph Contrastive Learning ABSTRACT: Self-supervised learning has become a key method for training deep learning models when labeled data is scarce or unavailable. While graph machine learning holds great promise across various domains, the design of effective pretext tasks for self-supervised graph representation learning remains challenging. Contrastive learning, a popular approach in graph self-supervised learning, leverages positive and negative pairs to compute a contrastive loss function. However, current graph contrastive learning methods often struggle to fully use structural patterns and node similarities. To address these issues, we present a new method called Fused Gromov Wasserstein Subgraph Contrastive Learning (FOSSIL). Our model integrates node-level and subgraph-level contrastive learning, seamlessly combining a standard node-level contrastive loss with the Fused Gromov-Wasserstein distance. This combination helps our method capture both node features and graph structure together. Importantly, our approach works well with both homophilic and heterophilic graphs and can dynamically create views for generating positive and negative pairs. Through extensive experiments on benchmark graph datasets, we show that FOSSIL outperforms or achieves competitive performance compared to current state-of-the-art methods.
no_new_dataset
0.949012
2502.20897
Matthias Orlikowski
Matthias Orlikowski, Jiaxin Pei, Paul R\"ottger, Philipp Cimiano, David Jurgens, Dirk Hovy
Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals' Subjective Text Perceptions
Reviewed ARR December 2024
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
People naturally vary in their annotations for subjective questions and some of this variation is thought to be due to the person's sociodemographic characteristics. LLMs have also been used to label data, but recent work has shown that models perform poorly when prompted with sociodemographic attributes, suggesting limited inherent sociodemographic knowledge. Here, we ask whether LLMs can be trained to be accurate sociodemographic models of annotator variation. Using a curated dataset of five tasks with standardized sociodemographics, we show that models do improve in sociodemographic prompting when trained but that this performance gain is largely due to models learning annotator-specific behaviour rather than sociodemographic patterns. Across all tasks, our results suggest that models learn little meaningful connection between sociodemographics and annotation, raising doubts about the current use of LLMs for simulating sociodemographic variation and behaviour.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 09:53:42 GMT" } ]
2025-03-03T00:00:00
[ [ "Orlikowski", "Matthias", "" ], [ "Pei", "Jiaxin", "" ], [ "Röttger", "Paul", "" ], [ "Cimiano", "Philipp", "" ], [ "Jurgens", "David", "" ], [ "Hovy", "Dirk", "" ] ]
TITLE: Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals' Subjective Text Perceptions ABSTRACT: People naturally vary in their annotations for subjective questions and some of this variation is thought to be due to the person's sociodemographic characteristics. LLMs have also been used to label data, but recent work has shown that models perform poorly when prompted with sociodemographic attributes, suggesting limited inherent sociodemographic knowledge. Here, we ask whether LLMs can be trained to be accurate sociodemographic models of annotator variation. Using a curated dataset of five tasks with standardized sociodemographics, we show that models do improve in sociodemographic prompting when trained but that this performance gain is largely due to models learning annotator-specific behaviour rather than sociodemographic patterns. Across all tasks, our results suggest that models learn little meaningful connection between sociodemographics and annotation, raising doubts about the current use of LLMs for simulating sociodemographic variation and behaviour.
new_dataset
0.950732
2502.20925
Nu Hoang
Bao Duong, Nu Hoang, Thin Nguyen
Amortized Conditional Independence Testing
Accepted at PAKDD 2025
null
null
null
stat.ML cs.LG
http://creativecommons.org/licenses/by/4.0/
Testing for the conditional independence structure in data is a fundamental and critical task in statistics and machine learning, which finds natural applications in causal discovery - a highly relevant problem to many scientific disciplines. Existing methods seek to design explicit test statistics that quantify the degree of conditional dependence, which is highly challenging yet cannot capture nor utilize prior knowledge in a data-driven manner. In this study, an entirely new approach is introduced, where we instead propose to amortize conditional independence testing and devise ACID - a novel transformer-based neural network architecture that learns to test for conditional independence. ACID can be trained on synthetic data in a supervised learning fashion, and the learned model can then be applied to any dataset of similar natures or adapted to new domains by fine-tuning with a negligible computational cost. Our extensive empirical evaluations on both synthetic and real data reveal that ACID consistently achieves state-of-the-art performance against existing baselines under multiple metrics, and is able to generalize robustly to unseen sample sizes, dimensionalities, as well as non-linearities with a remarkably low inference time.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 10:29:56 GMT" } ]
2025-03-03T00:00:00
[ [ "Duong", "Bao", "" ], [ "Hoang", "Nu", "" ], [ "Nguyen", "Thin", "" ] ]
TITLE: Amortized Conditional Independence Testing ABSTRACT: Testing for the conditional independence structure in data is a fundamental and critical task in statistics and machine learning, which finds natural applications in causal discovery - a highly relevant problem to many scientific disciplines. Existing methods seek to design explicit test statistics that quantify the degree of conditional dependence, which is highly challenging yet cannot capture nor utilize prior knowledge in a data-driven manner. In this study, an entirely new approach is introduced, where we instead propose to amortize conditional independence testing and devise ACID - a novel transformer-based neural network architecture that learns to test for conditional independence. ACID can be trained on synthetic data in a supervised learning fashion, and the learned model can then be applied to any dataset of similar natures or adapted to new domains by fine-tuning with a negligible computational cost. Our extensive empirical evaluations on both synthetic and real data reveal that ACID consistently achieves state-of-the-art performance against existing baselines under multiple metrics, and is able to generalize robustly to unseen sample sizes, dimensionalities, as well as non-linearities with a remarkably low inference time.
no_new_dataset
0.947962
2502.20931
Ilya Koziev
Ilya Koziev
Automated Evaluation of Meter and Rhyme in Russian Generative and Human-Authored Poetry
7 pages, 1 figure
null
null
null
cs.CL
http://creativecommons.org/licenses/by-nc-sa/4.0/
Generative poetry systems require effective tools for data engineering and automatic evaluation, particularly to assess how well a poem adheres to versification rules, such as the correct alternation of stressed and unstressed syllables and the presence of rhymes. In this work, we introduce the Russian Poetry Scansion Tool library designed for stress mark placement in Russian-language syllabo-tonic poetry, rhyme detection, and identification of defects of poeticness. Additionally, we release RIFMA -- a dataset of poem fragments spanning various genres and forms, annotated with stress marks. This dataset can be used to evaluate the capability of modern large language models to accurately place stress marks in poetic texts. The published resources provide valuable tools for researchers and practitioners in the field of creative generative AI, facilitating advancements in the development and evaluation of generative poetry systems.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 10:39:07 GMT" } ]
2025-03-03T00:00:00
[ [ "Koziev", "Ilya", "" ] ]
TITLE: Automated Evaluation of Meter and Rhyme in Russian Generative and Human-Authored Poetry ABSTRACT: Generative poetry systems require effective tools for data engineering and automatic evaluation, particularly to assess how well a poem adheres to versification rules, such as the correct alternation of stressed and unstressed syllables and the presence of rhymes. In this work, we introduce the Russian Poetry Scansion Tool library designed for stress mark placement in Russian-language syllabo-tonic poetry, rhyme detection, and identification of defects of poeticness. Additionally, we release RIFMA -- a dataset of poem fragments spanning various genres and forms, annotated with stress marks. This dataset can be used to evaluate the capability of modern large language models to accurately place stress marks in poetic texts. The published resources provide valuable tools for researchers and practitioners in the field of creative generative AI, facilitating advancements in the development and evaluation of generative poetry systems.
new_dataset
0.960473
2502.20936
Michael Dinzinger
Michael Dinzinger and Laura Caspari and Kanishka Ghosh Dastidar and Jelena Mitrovi\'c and Michael Granitzer
WebFAQ: A Multilingual Collection of Natural Q&A Datasets for Dense Retrieval
10 pages, 3 figures, 7 tables
null
null
null
cs.CL cs.AI cs.IR
http://creativecommons.org/licenses/by/4.0/
We present WebFAQ, a large-scale collection of open-domain question answering datasets derived from FAQ-style schema.org annotations. In total, the data collection consists of 96 million natural question-answer (QA) pairs across 75 languages, including 47 million (49%) non-English samples. WebFAQ further serves as the foundation for 20 monolingual retrieval benchmarks with a total size of 11.2 million QA pairs (5.9 million non-English). These datasets are carefully curated through refined filtering and near-duplicate detection, yielding high-quality resources for training and evaluating multilingual dense retrieval models. To empirically confirm WebFAQ's efficacy, we use the collected QAs to fine-tune an in-domain pretrained XLM-RoBERTa model. Through this process of dataset-specific fine-tuning, the model achieves significant retrieval performance gains, which generalize - beyond WebFAQ - to other multilingual retrieval benchmarks evaluated in zero-shot setting. Last but not least, we utilize WebFAQ to construct a set of QA-aligned bilingual corpora spanning over 1000 language pairs using state-of-the-art bitext mining and automated LLM-assessed translation evaluation. Due to our advanced, automated method of bitext dataset generation, the resulting bilingual corpora demonstrate higher translation quality compared to similar datasets. WebFAQ and all associated resources are publicly available on GitHub and HuggingFace.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 10:46:52 GMT" } ]
2025-03-03T00:00:00
[ [ "Dinzinger", "Michael", "" ], [ "Caspari", "Laura", "" ], [ "Dastidar", "Kanishka Ghosh", "" ], [ "Mitrović", "Jelena", "" ], [ "Granitzer", "Michael", "" ] ]
TITLE: WebFAQ: A Multilingual Collection of Natural Q&A Datasets for Dense Retrieval ABSTRACT: We present WebFAQ, a large-scale collection of open-domain question answering datasets derived from FAQ-style schema.org annotations. In total, the data collection consists of 96 million natural question-answer (QA) pairs across 75 languages, including 47 million (49%) non-English samples. WebFAQ further serves as the foundation for 20 monolingual retrieval benchmarks with a total size of 11.2 million QA pairs (5.9 million non-English). These datasets are carefully curated through refined filtering and near-duplicate detection, yielding high-quality resources for training and evaluating multilingual dense retrieval models. To empirically confirm WebFAQ's efficacy, we use the collected QAs to fine-tune an in-domain pretrained XLM-RoBERTa model. Through this process of dataset-specific fine-tuning, the model achieves significant retrieval performance gains, which generalize - beyond WebFAQ - to other multilingual retrieval benchmarks evaluated in zero-shot setting. Last but not least, we utilize WebFAQ to construct a set of QA-aligned bilingual corpora spanning over 1000 language pairs using state-of-the-art bitext mining and automated LLM-assessed translation evaluation. Due to our advanced, automated method of bitext dataset generation, the resulting bilingual corpora demonstrate higher translation quality compared to similar datasets. WebFAQ and all associated resources are publicly available on GitHub and HuggingFace.
new_dataset
0.519312
2502.20948
Petr Sokerin Mr
Petr Sokerin, Dmitry Anikin, Sofia Krehova, Alexey Zaytsev
Concealed Adversarial attacks on neural networks for sequential data
null
null
null
null
cs.LG cs.AI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The emergence of deep learning led to the broad usage of neural networks in the time series domain for various applications, including finance and medicine. While powerful, these models are prone to adversarial attacks: a benign targeted perturbation of input data leads to significant changes in a classifier's output. However, formally small attacks in the time series domain become easily detected by the human eye or a simple detector model. We develop a concealed adversarial attack for different time-series models: it provides more realistic perturbations, being hard to detect by a human or model discriminator. To achieve this goal, the proposed adversarial attack maximizes an aggregation of a classifier and a trained discriminator loss. To make the attack stronger, we also propose a training procedure for a discriminator that provides broader coverage of possible attacks. Extensive benchmarking on six UCR time series datasets across four diverse architectures - including recurrent, convolutional, state-space, and transformer-based models - demonstrates the superiority of our attack for a concealability-efficiency trade-off. Our findings highlight the growing challenge of designing robust time series models, emphasizing the need for improved defenses against realistic and effective attacks.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:03:32 GMT" } ]
2025-03-03T00:00:00
[ [ "Sokerin", "Petr", "" ], [ "Anikin", "Dmitry", "" ], [ "Krehova", "Sofia", "" ], [ "Zaytsev", "Alexey", "" ] ]
TITLE: Concealed Adversarial attacks on neural networks for sequential data ABSTRACT: The emergence of deep learning led to the broad usage of neural networks in the time series domain for various applications, including finance and medicine. While powerful, these models are prone to adversarial attacks: a benign targeted perturbation of input data leads to significant changes in a classifier's output. However, formally small attacks in the time series domain become easily detected by the human eye or a simple detector model. We develop a concealed adversarial attack for different time-series models: it provides more realistic perturbations, being hard to detect by a human or model discriminator. To achieve this goal, the proposed adversarial attack maximizes an aggregation of a classifier and a trained discriminator loss. To make the attack stronger, we also propose a training procedure for a discriminator that provides broader coverage of possible attacks. Extensive benchmarking on six UCR time series datasets across four diverse architectures - including recurrent, convolutional, state-space, and transformer-based models - demonstrates the superiority of our attack for a concealability-efficiency trade-off. Our findings highlight the growing challenge of designing robust time series models, emphasizing the need for improved defenses against realistic and effective attacks.
no_new_dataset
0.945349
2502.20954
Jindong Li
Jindong Li and Tim Hamann and Jens Barth and Peter Kaempf and Dario Zanca and Bjoern Eskofier
Robust and Efficient Writer-Independent IMU-Based Handwriting Recognization
null
null
null
null
cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Online handwriting recognition (HWR) using data from inertial measurement units (IMUs) remains challenging due to variations in writing styles and the limited availability of high-quality annotated datasets. Traditional models often struggle to recognize handwriting from unseen writers, making writer-independent (WI) recognition a crucial but difficult problem. This paper presents an HWR model with an encoder-decoder structure for IMU data, featuring a CNN-based encoder for feature extraction and a BiLSTM decoder for sequence modeling, which supports inputs of varying lengths. Our approach demonstrates strong robustness and data efficiency, outperforming existing methods on WI datasets, including the WI split of the OnHW dataset and our own dataset. Extensive evaluations show that our model maintains high accuracy across different age groups and writing conditions while effectively learning from limited data. Through comprehensive ablation studies, we analyze key design choices, achieving a balance between accuracy and efficiency. These findings contribute to the development of more adaptable and scalable HWR systems for real-world applications.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:09:28 GMT" } ]
2025-03-03T00:00:00
[ [ "Li", "Jindong", "" ], [ "Hamann", "Tim", "" ], [ "Barth", "Jens", "" ], [ "Kaempf", "Peter", "" ], [ "Zanca", "Dario", "" ], [ "Eskofier", "Bjoern", "" ] ]
TITLE: Robust and Efficient Writer-Independent IMU-Based Handwriting Recognization ABSTRACT: Online handwriting recognition (HWR) using data from inertial measurement units (IMUs) remains challenging due to variations in writing styles and the limited availability of high-quality annotated datasets. Traditional models often struggle to recognize handwriting from unseen writers, making writer-independent (WI) recognition a crucial but difficult problem. This paper presents an HWR model with an encoder-decoder structure for IMU data, featuring a CNN-based encoder for feature extraction and a BiLSTM decoder for sequence modeling, which supports inputs of varying lengths. Our approach demonstrates strong robustness and data efficiency, outperforming existing methods on WI datasets, including the WI split of the OnHW dataset and our own dataset. Extensive evaluations show that our model maintains high accuracy across different age groups and writing conditions while effectively learning from limited data. Through comprehensive ablation studies, we analyze key design choices, achieving a balance between accuracy and efficiency. These findings contribute to the development of more adaptable and scalable HWR systems for real-world applications.
new_dataset
0.966663
2502.20957
Giseung Park
Giseung Park, Youngchul Sung
Reward Dimension Reduction for Scalable Multi-Objective Reinforcement Learning
Accepted to ICLR 2025
null
null
null
cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In this paper, we introduce a simple yet effective reward dimension reduction method to tackle the scalability challenges of multi-objective reinforcement learning algorithms. While most existing approaches focus on optimizing two to four objectives, their abilities to scale to environments with more objectives remain uncertain. Our method uses a dimension reduction approach to enhance learning efficiency and policy performance in multi-objective settings. While most traditional dimension reduction methods are designed for static datasets, our approach is tailored for online learning and preserves Pareto-optimality after transformation. We propose a new training and evaluation framework for reward dimension reduction in multi-objective reinforcement learning and demonstrate the superiority of our method in environments including one with sixteen objectives, significantly outperforming existing online dimension reduction methods.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:13:23 GMT" } ]
2025-03-03T00:00:00
[ [ "Park", "Giseung", "" ], [ "Sung", "Youngchul", "" ] ]
TITLE: Reward Dimension Reduction for Scalable Multi-Objective Reinforcement Learning ABSTRACT: In this paper, we introduce a simple yet effective reward dimension reduction method to tackle the scalability challenges of multi-objective reinforcement learning algorithms. While most existing approaches focus on optimizing two to four objectives, their abilities to scale to environments with more objectives remain uncertain. Our method uses a dimension reduction approach to enhance learning efficiency and policy performance in multi-objective settings. While most traditional dimension reduction methods are designed for static datasets, our approach is tailored for online learning and preserves Pareto-optimality after transformation. We propose a new training and evaluation framework for reward dimension reduction in multi-objective reinforcement learning and demonstrate the superiority of our method in environments including one with sixteen objectives, significantly outperforming existing online dimension reduction methods.
no_new_dataset
0.946101
2502.20966
Richard Scott Bergna
Richard Bergna, Stefan Depeweg, Sergio Calvo Ordonez, Jonathan Plenk, Alvaro Cartea, Jose Miguel Hernandez-Lobato
Post-Hoc Uncertainty Quantification in Pre-Trained Neural Networks via Activation-Level Gaussian Processes
10 pages, 8 figures, 7th Symposium on Advances in Approximate Bayesian Inference
null
null
null
stat.ML cs.LG
http://creativecommons.org/licenses/by/4.0/
Uncertainty quantification in neural networks through methods such as Dropout, Bayesian neural networks and Laplace approximations is either prone to underfitting or computationally demanding, rendering these approaches impractical for large-scale datasets. In this work, we address these shortcomings by shifting the focus from uncertainty in the weight space to uncertainty at the activation level, via Gaussian processes. More specifically, we introduce the Gaussian Process Activation function (GAPA) to capture neuron-level uncertainties. Our approach operates in a post-hoc manner, preserving the original mean predictions of the pre-trained neural network and thereby avoiding the underfitting issues commonly encountered in previous methods. We propose two methods. The first, GAPA-Free, employs empirical kernel learning from the training data for the hyperparameters and is highly efficient during training. The second, GAPA-Variational, learns the hyperparameters via gradient descent on the kernels, thus affording greater flexibility. Empirical results demonstrate that GAPA-Variational outperforms the Laplace approximation on most datasets in at least one of the uncertainty quantification metrics.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:29:06 GMT" } ]
2025-03-03T00:00:00
[ [ "Bergna", "Richard", "" ], [ "Depeweg", "Stefan", "" ], [ "Ordonez", "Sergio Calvo", "" ], [ "Plenk", "Jonathan", "" ], [ "Cartea", "Alvaro", "" ], [ "Hernandez-Lobato", "Jose Miguel", "" ] ]
TITLE: Post-Hoc Uncertainty Quantification in Pre-Trained Neural Networks via Activation-Level Gaussian Processes ABSTRACT: Uncertainty quantification in neural networks through methods such as Dropout, Bayesian neural networks and Laplace approximations is either prone to underfitting or computationally demanding, rendering these approaches impractical for large-scale datasets. In this work, we address these shortcomings by shifting the focus from uncertainty in the weight space to uncertainty at the activation level, via Gaussian processes. More specifically, we introduce the Gaussian Process Activation function (GAPA) to capture neuron-level uncertainties. Our approach operates in a post-hoc manner, preserving the original mean predictions of the pre-trained neural network and thereby avoiding the underfitting issues commonly encountered in previous methods. We propose two methods. The first, GAPA-Free, employs empirical kernel learning from the training data for the hyperparameters and is highly efficient during training. The second, GAPA-Variational, learns the hyperparameters via gradient descent on the kernels, thus affording greater flexibility. Empirical results demonstrate that GAPA-Variational outperforms the Laplace approximation on most datasets in at least one of the uncertainty quantification metrics.
no_new_dataset
0.952926
2502.20975
Naman Bansal
Naman Bansal, Yash mahajan, Sanjeev Sinha, Santu Karmaker
Set-Theoretic Compositionality of Sentence Embeddings
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
Sentence encoders play a pivotal role in various NLP tasks; hence, an accurate evaluation of their compositional properties is paramount. However, existing evaluation methods predominantly focus on goal task-specific performance. This leaves a significant gap in understanding how well sentence embeddings demonstrate fundamental compositional properties in a task-independent context. Leveraging classical set theory, we address this gap by proposing six criteria based on three core "set-like" compositions/operations: \textit{TextOverlap}, \textit{TextDifference}, and \textit{TextUnion}. We systematically evaluate $7$ classical and $9$ Large Language Model (LLM)-based sentence encoders to assess their alignment with these criteria. Our findings show that SBERT consistently demonstrates set-like compositional properties, surpassing even the latest LLMs. Additionally, we introduce a new dataset of ~$192$K samples designed to facilitate future benchmarking efforts on set-like compositionality of sentence embeddings.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:40:34 GMT" } ]
2025-03-03T00:00:00
[ [ "Bansal", "Naman", "" ], [ "mahajan", "Yash", "" ], [ "Sinha", "Sanjeev", "" ], [ "Karmaker", "Santu", "" ] ]
TITLE: Set-Theoretic Compositionality of Sentence Embeddings ABSTRACT: Sentence encoders play a pivotal role in various NLP tasks; hence, an accurate evaluation of their compositional properties is paramount. However, existing evaluation methods predominantly focus on goal task-specific performance. This leaves a significant gap in understanding how well sentence embeddings demonstrate fundamental compositional properties in a task-independent context. Leveraging classical set theory, we address this gap by proposing six criteria based on three core "set-like" compositions/operations: \textit{TextOverlap}, \textit{TextDifference}, and \textit{TextUnion}. We systematically evaluate $7$ classical and $9$ Large Language Model (LLM)-based sentence encoders to assess their alignment with these criteria. Our findings show that SBERT consistently demonstrates set-like compositional properties, surpassing even the latest LLMs. Additionally, we introduce a new dataset of ~$192$K samples designed to facilitate future benchmarking efforts on set-like compositionality of sentence embeddings.
new_dataset
0.963265
2502.20981
Fuyun Wang
Fuyun Wang, Tong Zhang, Yuanzhi Wang, Yide Qiu, Xin Liu, Xu Guo, Zhen Cui
Distribution Prototype Diffusion Learning for Open-set Supervised Anomaly Detection
Accepted by CVPR 2025
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In Open-set Supervised Anomaly Detection (OSAD), the existing methods typically generate pseudo anomalies to compensate for the scarcity of observed anomaly samples, while overlooking critical priors of normal samples, leading to less effective discriminative boundaries. To address this issue, we propose a Distribution Prototype Diffusion Learning (DPDL) method aimed at enclosing normal samples within a compact and discriminative distribution space. Specifically, we construct multiple learnable Gaussian prototypes to create a latent representation space for abundant and diverse normal samples and learn a Schr\"odinger bridge to facilitate a diffusive transition toward these prototypes for normal samples while steering anomaly samples away. Moreover, to enhance inter-sample separation, we design a dispersion feature learning way in hyperspherical space, which benefits the identification of out-of-distribution anomalies. Experimental results demonstrate the effectiveness and superiority of our proposed DPDL, achieving state-of-the-art performance on 9 public datasets.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:50:50 GMT" } ]
2025-03-03T00:00:00
[ [ "Wang", "Fuyun", "" ], [ "Zhang", "Tong", "" ], [ "Wang", "Yuanzhi", "" ], [ "Qiu", "Yide", "" ], [ "Liu", "Xin", "" ], [ "Guo", "Xu", "" ], [ "Cui", "Zhen", "" ] ]
TITLE: Distribution Prototype Diffusion Learning for Open-set Supervised Anomaly Detection ABSTRACT: In Open-set Supervised Anomaly Detection (OSAD), the existing methods typically generate pseudo anomalies to compensate for the scarcity of observed anomaly samples, while overlooking critical priors of normal samples, leading to less effective discriminative boundaries. To address this issue, we propose a Distribution Prototype Diffusion Learning (DPDL) method aimed at enclosing normal samples within a compact and discriminative distribution space. Specifically, we construct multiple learnable Gaussian prototypes to create a latent representation space for abundant and diverse normal samples and learn a Schr\"odinger bridge to facilitate a diffusive transition toward these prototypes for normal samples while steering anomaly samples away. Moreover, to enhance inter-sample separation, we design a dispersion feature learning way in hyperspherical space, which benefits the identification of out-of-distribution anomalies. Experimental results demonstrate the effectiveness and superiority of our proposed DPDL, achieving state-of-the-art performance on 9 public datasets.
no_new_dataset
0.954052
2502.20985
Maximilian Rouven Rokuss
Maximilian Rokuss, Yannick Kirchhoff, Seval Akbal, Balint Kovacs, Saikat Roy, Constantin Ulrich, Tassilo Wald, Lukas T. Rotkopf, Heinz-Peter Schlemmer, Klaus Maier-Hein
LesionLocator: Zero-Shot Universal Tumor Segmentation and Tracking in 3D Whole-Body Imaging
Accepted at CVPR 2025
null
null
null
cs.CV cs.AI
http://creativecommons.org/licenses/by/4.0/
In this work, we present LesionLocator, a framework for zero-shot longitudinal lesion tracking and segmentation in 3D medical imaging, establishing the first end-to-end model capable of 4D tracking with dense spatial prompts. Our model leverages an extensive dataset of 23,262 annotated medical scans, as well as synthesized longitudinal data across diverse lesion types. The diversity and scale of our dataset significantly enhances model generalizability to real-world medical imaging challenges and addresses key limitations in longitudinal data availability. LesionLocator outperforms all existing promptable models in lesion segmentation by nearly 10 dice points, reaching human-level performance, and achieves state-of-the-art results in lesion tracking, with superior lesion retrieval and segmentation accuracy. LesionLocator not only sets a new benchmark in universal promptable lesion segmentation and automated longitudinal lesion tracking but also provides the first open-access solution of its kind, releasing our synthetic 4D dataset and model to the community, empowering future advancements in medical imaging. Code is available at: www.github.com/MIC-DKFZ/LesionLocator
[ { "version": "v1", "created": "Fri, 28 Feb 2025 11:58:33 GMT" } ]
2025-03-03T00:00:00
[ [ "Rokuss", "Maximilian", "" ], [ "Kirchhoff", "Yannick", "" ], [ "Akbal", "Seval", "" ], [ "Kovacs", "Balint", "" ], [ "Roy", "Saikat", "" ], [ "Ulrich", "Constantin", "" ], [ "Wald", "Tassilo", "" ], [ "Rotkopf", "Lukas T.", "" ], [ "Schlemmer", "Heinz-Peter", "" ], [ "Maier-Hein", "Klaus", "" ] ]
TITLE: LesionLocator: Zero-Shot Universal Tumor Segmentation and Tracking in 3D Whole-Body Imaging ABSTRACT: In this work, we present LesionLocator, a framework for zero-shot longitudinal lesion tracking and segmentation in 3D medical imaging, establishing the first end-to-end model capable of 4D tracking with dense spatial prompts. Our model leverages an extensive dataset of 23,262 annotated medical scans, as well as synthesized longitudinal data across diverse lesion types. The diversity and scale of our dataset significantly enhances model generalizability to real-world medical imaging challenges and addresses key limitations in longitudinal data availability. LesionLocator outperforms all existing promptable models in lesion segmentation by nearly 10 dice points, reaching human-level performance, and achieves state-of-the-art results in lesion tracking, with superior lesion retrieval and segmentation accuracy. LesionLocator not only sets a new benchmark in universal promptable lesion segmentation and automated longitudinal lesion tracking but also provides the first open-access solution of its kind, releasing our synthetic 4D dataset and model to the community, empowering future advancements in medical imaging. Code is available at: www.github.com/MIC-DKFZ/LesionLocator
new_dataset
0.941277
2502.20988
Jinguang Gu
Qiyuan Li, Haijiang Liu, Caicai Guo, Deyu Chen, Meng Wang, Feng Gao, Jinguang Gu
Merging Clinical Knowledge into Large Language Models for Medical Research and Applications: A Survey
null
null
null
null
cs.AI cs.CL
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Clinical knowledge is the collection of information learned from studies on the causes, prognosis, diagnosis, and treatment of diseases. This type of knowledge can improve curing performances, and promote physical health. With the emergence of large language models (LLMs), medical artificial intelligence (medical AI), which aims to apply academic medical AI systems to real-world medical scenarios, has entered a new age of development, resulting in excellent works such as DoctorGPT and Pangu-Drug from academic and industrial researches. However, the field lacks a comprehensive compendium and comparison of building medical AI systems from academia and industry. Therefore, this survey focuses on the building paradigms of medical AI systems including the use of clinical databases, datasets, training pipelines, integrating medical knowledge graphs, system applications, and evaluation systems. We hope that this survey can help relevant practical researchers understand the current performance of academic models in various fields of healthcare, as well as the potential problems and future directions for implementing these scientific achievements.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 12:00:51 GMT" } ]
2025-03-03T00:00:00
[ [ "Li", "Qiyuan", "" ], [ "Liu", "Haijiang", "" ], [ "Guo", "Caicai", "" ], [ "Chen", "Deyu", "" ], [ "Wang", "Meng", "" ], [ "Gao", "Feng", "" ], [ "Gu", "Jinguang", "" ] ]
TITLE: Merging Clinical Knowledge into Large Language Models for Medical Research and Applications: A Survey ABSTRACT: Clinical knowledge is the collection of information learned from studies on the causes, prognosis, diagnosis, and treatment of diseases. This type of knowledge can improve curing performances, and promote physical health. With the emergence of large language models (LLMs), medical artificial intelligence (medical AI), which aims to apply academic medical AI systems to real-world medical scenarios, has entered a new age of development, resulting in excellent works such as DoctorGPT and Pangu-Drug from academic and industrial researches. However, the field lacks a comprehensive compendium and comparison of building medical AI systems from academia and industry. Therefore, this survey focuses on the building paradigms of medical AI systems including the use of clinical databases, datasets, training pipelines, integrating medical knowledge graphs, system applications, and evaluation systems. We hope that this survey can help relevant practical researchers understand the current performance of academic models in various fields of healthcare, as well as the potential problems and future directions for implementing these scientific achievements.
no_new_dataset
0.941331
2502.20992
Jiaxiang Liu
Xiusheng Huang, Jiaxiang Liu, Yequan Wang, Jun Zhao, Kang Liu
Capability Localization: Capabilities Can be Localized rather than Individual Knowledge
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by-nc-nd/4.0/
Large scale language models have achieved superior performance in tasks related to natural language processing, however, it is still unclear how model parameters affect performance improvement. Previous studies assumed that individual knowledge is stored in local parameters, and the storage form of individual knowledge is dispersed parameters, parameter layers, or parameter chains, which are not unified. We found through fidelity and reliability evaluation experiments that individual knowledge cannot be localized. Afterwards, we constructed a dataset for decoupling experiments and discovered the potential for localizing data commonalities. To further reveal this phenomenon, this paper proposes a Commonality Neuron Localization (CNL) method, which successfully locates commonality neurons and achieves a neuron overlap rate of 96.42% on the GSM8K dataset. Finally, we have demonstrated through cross data experiments that commonality neurons are a collection of capability neurons that possess the capability to enhance performance. Our code is available at https://github.com/nlpkeg/Capability-Neuron-Localization.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 12:22:13 GMT" } ]
2025-03-03T00:00:00
[ [ "Huang", "Xiusheng", "" ], [ "Liu", "Jiaxiang", "" ], [ "Wang", "Yequan", "" ], [ "Zhao", "Jun", "" ], [ "Liu", "Kang", "" ] ]
TITLE: Capability Localization: Capabilities Can be Localized rather than Individual Knowledge ABSTRACT: Large scale language models have achieved superior performance in tasks related to natural language processing, however, it is still unclear how model parameters affect performance improvement. Previous studies assumed that individual knowledge is stored in local parameters, and the storage form of individual knowledge is dispersed parameters, parameter layers, or parameter chains, which are not unified. We found through fidelity and reliability evaluation experiments that individual knowledge cannot be localized. Afterwards, we constructed a dataset for decoupling experiments and discovered the potential for localizing data commonalities. To further reveal this phenomenon, this paper proposes a Commonality Neuron Localization (CNL) method, which successfully locates commonality neurons and achieves a neuron overlap rate of 96.42% on the GSM8K dataset. Finally, we have demonstrated through cross data experiments that commonality neurons are a collection of capability neurons that possess the capability to enhance performance. Our code is available at https://github.com/nlpkeg/Capability-Neuron-Localization.
new_dataset
0.962673
2502.21011
Junchao Zhu
Junchao Zhu, Ruining Deng, Tianyuan Yao, Juming Xiong, Chongyu Qu, Junlin Guo, Siqi Lu, Yucheng Tang, Daguang Xu, Mengmeng Yin, Yu Wang, Shilin Zhao, Yaohong Wang, Haichun Yang, Yuankai Huo
MagNet: Multi-Level Attention Graph Network for Predicting High-Resolution Spatial Transcriptomics
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The rapid development of spatial transcriptomics (ST) offers new opportunities to explore the gene expression patterns within the spatial microenvironment. Current research integrates pathological images to infer gene expression, addressing the high costs and time-consuming processes to generate spatial transcriptomics data. However, as spatial transcriptomics resolution continues to improve, existing methods remain primarily focused on gene expression prediction at low-resolution spot levels. These methods face significant challenges, especially the information bottleneck, when they are applied to high-resolution HD data. To bridge this gap, this paper introduces MagNet, a multi-level attention graph network designed for accurate prediction of high-resolution HD data. MagNet employs cross-attention layers to integrate features from multi-resolution image patches hierarchically and utilizes a GAT-Transformer module to aggregate neighborhood information. By integrating multilevel features, MagNet overcomes the limitations posed by low-resolution inputs in predicting high-resolution gene expression. We systematically evaluated MagNet and existing ST prediction models on both a private spatial transcriptomics dataset and a public dataset at three different resolution levels. The results demonstrate that MagNet achieves state-of-the-art performance at both spot level and high-resolution bin levels, providing a novel methodology and benchmark for future research and applications in high-resolution HD-level spatial transcriptomics. Code is available at https://github.com/Junchao-Zhu/MagNet.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 12:55:37 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhu", "Junchao", "" ], [ "Deng", "Ruining", "" ], [ "Yao", "Tianyuan", "" ], [ "Xiong", "Juming", "" ], [ "Qu", "Chongyu", "" ], [ "Guo", "Junlin", "" ], [ "Lu", "Siqi", "" ], [ "Tang", "Yucheng", "" ], [ "Xu", "Daguang", "" ], [ "Yin", "Mengmeng", "" ], [ "Wang", "Yu", "" ], [ "Zhao", "Shilin", "" ], [ "Wang", "Yaohong", "" ], [ "Yang", "Haichun", "" ], [ "Huo", "Yuankai", "" ] ]
TITLE: MagNet: Multi-Level Attention Graph Network for Predicting High-Resolution Spatial Transcriptomics ABSTRACT: The rapid development of spatial transcriptomics (ST) offers new opportunities to explore the gene expression patterns within the spatial microenvironment. Current research integrates pathological images to infer gene expression, addressing the high costs and time-consuming processes to generate spatial transcriptomics data. However, as spatial transcriptomics resolution continues to improve, existing methods remain primarily focused on gene expression prediction at low-resolution spot levels. These methods face significant challenges, especially the information bottleneck, when they are applied to high-resolution HD data. To bridge this gap, this paper introduces MagNet, a multi-level attention graph network designed for accurate prediction of high-resolution HD data. MagNet employs cross-attention layers to integrate features from multi-resolution image patches hierarchically and utilizes a GAT-Transformer module to aggregate neighborhood information. By integrating multilevel features, MagNet overcomes the limitations posed by low-resolution inputs in predicting high-resolution gene expression. We systematically evaluated MagNet and existing ST prediction models on both a private spatial transcriptomics dataset and a public dataset at three different resolution levels. The results demonstrate that MagNet achieves state-of-the-art performance at both spot level and high-resolution bin levels, providing a novel methodology and benchmark for future research and applications in high-resolution HD-level spatial transcriptomics. Code is available at https://github.com/Junchao-Zhu/MagNet.
no_new_dataset
0.952882
2502.21012
Silin Chen
Silin Chen, Kangjian Di, Yichu Xu, Han-Jia Ye, Wenhan Luo, Ningmu Zou
FedDyMem: Efficient Federated Learning with Dynamic Memory and Memory-Reduce for Unsupervised Image Anomaly Detection
null
null
null
null
cs.DC cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Unsupervised image anomaly detection (UAD) has become a critical process in industrial and medical applications, but it faces growing challenges due to increasing concerns over data privacy. The limited class diversity inherent to one-class classification tasks, combined with distribution biases caused by variations in products across and within clients, poses significant challenges for preserving data privacy with federated UAD. Thus, this article proposes an efficient federated learning method with dynamic memory and memory-reduce for unsupervised image anomaly detection, called FedDyMem. Considering all client data belongs to a single class (i.e., normal sample) in UAD and the distribution of intra-class features demonstrates significant skewness, FedDyMem facilitates knowledge sharing between the client and server through the client's dynamic memory bank instead of model parameters. In the local clients, a memory generator and a metric loss are employed to improve the consistency of the feature distribution for normal samples, leveraging the local model to update the memory bank dynamically. For efficient communication, a memory-reduce method based on weighted averages is proposed to significantly decrease the scale of memory banks. On the server, global memory is constructed and distributed to individual clients through k-means aggregation. Experiments conducted on six industrial and medical datasets, comprising a mixture of six products or health screening types derived from eleven public datasets, demonstrate the effectiveness of FedDyMem.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 12:55:58 GMT" } ]
2025-03-03T00:00:00
[ [ "Chen", "Silin", "" ], [ "Di", "Kangjian", "" ], [ "Xu", "Yichu", "" ], [ "Ye", "Han-Jia", "" ], [ "Luo", "Wenhan", "" ], [ "Zou", "Ningmu", "" ] ]
TITLE: FedDyMem: Efficient Federated Learning with Dynamic Memory and Memory-Reduce for Unsupervised Image Anomaly Detection ABSTRACT: Unsupervised image anomaly detection (UAD) has become a critical process in industrial and medical applications, but it faces growing challenges due to increasing concerns over data privacy. The limited class diversity inherent to one-class classification tasks, combined with distribution biases caused by variations in products across and within clients, poses significant challenges for preserving data privacy with federated UAD. Thus, this article proposes an efficient federated learning method with dynamic memory and memory-reduce for unsupervised image anomaly detection, called FedDyMem. Considering all client data belongs to a single class (i.e., normal sample) in UAD and the distribution of intra-class features demonstrates significant skewness, FedDyMem facilitates knowledge sharing between the client and server through the client's dynamic memory bank instead of model parameters. In the local clients, a memory generator and a metric loss are employed to improve the consistency of the feature distribution for normal samples, leveraging the local model to update the memory bank dynamically. For efficient communication, a memory-reduce method based on weighted averages is proposed to significantly decrease the scale of memory banks. On the server, global memory is constructed and distributed to individual clients through k-means aggregation. Experiments conducted on six industrial and medical datasets, comprising a mixture of six products or health screening types derived from eleven public datasets, demonstrate the effectiveness of FedDyMem.
no_new_dataset
0.951863
2502.21034
Youran Zhou
Youran Zhou, Jianzhong Qi
Synthesizing Tabular Data Using Selectivity Enhanced Generative Adversarial Networks
This thesis submitted to the University of Melbourne for partial fulfillment of the degree of Master of Data Science
null
null
null
cs.LG cs.AI
http://creativecommons.org/licenses/by/4.0/
As E-commerce platforms face surging transactions during major shopping events like Black Friday, stress testing with synthesized data is crucial for resource planning. Most recent studies use Generative Adversarial Networks (GANs) to generate tabular data while ensuring privacy and machine learning utility. However, these methods overlook the computational demands of processing GAN-generated data, making them unsuitable for E-commerce stress testing. This thesis introduces a novel GAN-based approach incorporating query selectivity constraints, a key factor in database transaction processing. We integrate a pre-trained deep neural network to maintain selectivity consistency between real and synthetic data. Our method, tested on five real-world datasets, outperforms three state-of-the-art GANs and a VAE model, improving selectivity estimation accuracy by up to 20pct and machine learning utility by up to 6 pct.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:26:41 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhou", "Youran", "" ], [ "Qi", "Jianzhong", "" ] ]
TITLE: Synthesizing Tabular Data Using Selectivity Enhanced Generative Adversarial Networks ABSTRACT: As E-commerce platforms face surging transactions during major shopping events like Black Friday, stress testing with synthesized data is crucial for resource planning. Most recent studies use Generative Adversarial Networks (GANs) to generate tabular data while ensuring privacy and machine learning utility. However, these methods overlook the computational demands of processing GAN-generated data, making them unsuitable for E-commerce stress testing. This thesis introduces a novel GAN-based approach incorporating query selectivity constraints, a key factor in database transaction processing. We integrate a pre-trained deep neural network to maintain selectivity consistency between real and synthetic data. Our method, tested on five real-world datasets, outperforms three state-of-the-art GANs and a VAE model, improving selectivity estimation accuracy by up to 20pct and machine learning utility by up to 6 pct.
no_new_dataset
0.9455
2502.21035
Melanie Schaller Dr.
Melanie Schaller and Bodo Rosenhahn
S4ConvD: Adaptive Scaling and Frequency Adjustment for Energy-Efficient Sensor Networks in Smart Buildings
Submitted to TOSN Journal
null
null
null
cs.LG
http://creativecommons.org/licenses/by/4.0/
Predicting energy consumption in smart buildings is challenging due to dependencies in sensor data and the variability of environmental conditions. We introduce S4ConvD, a novel convolutional variant of Deep State Space Models (Deep-SSMs), that minimizes reliance on extensive preprocessing steps. S4ConvD is designed to optimize runtime in resource-constrained environments. By implementing adaptive scaling and frequency adjustments, this model shows to capture complex temporal patterns in building energy dynamics. Experiments on the ASHRAE Great Energy Predictor III dataset reveal that S4ConvD outperforms current benchmarks. Additionally, S4ConvD benefits from significant improvements in GPU runtime through the use of Block Tiling optimization techniques. Thus, S4ConvD has the potential for practical deployment in real-time energy modeling. Furthermore, the complete codebase and dataset are accessible on GitHub, fostering open-source contributions and facilitating further research. Our method also promotes resource-efficient model execution, enhancing both energy forecasting and the potential integration of renewable energy sources into smart grid systems.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:27:25 GMT" } ]
2025-03-03T00:00:00
[ [ "Schaller", "Melanie", "" ], [ "Rosenhahn", "Bodo", "" ] ]
TITLE: S4ConvD: Adaptive Scaling and Frequency Adjustment for Energy-Efficient Sensor Networks in Smart Buildings ABSTRACT: Predicting energy consumption in smart buildings is challenging due to dependencies in sensor data and the variability of environmental conditions. We introduce S4ConvD, a novel convolutional variant of Deep State Space Models (Deep-SSMs), that minimizes reliance on extensive preprocessing steps. S4ConvD is designed to optimize runtime in resource-constrained environments. By implementing adaptive scaling and frequency adjustments, this model shows to capture complex temporal patterns in building energy dynamics. Experiments on the ASHRAE Great Energy Predictor III dataset reveal that S4ConvD outperforms current benchmarks. Additionally, S4ConvD benefits from significant improvements in GPU runtime through the use of Block Tiling optimization techniques. Thus, S4ConvD has the potential for practical deployment in real-time energy modeling. Furthermore, the complete codebase and dataset are accessible on GitHub, fostering open-source contributions and facilitating further research. Our method also promotes resource-efficient model execution, enhancing both energy forecasting and the potential integration of renewable energy sources into smart grid systems.
no_new_dataset
0.947088
2502.21046
Jonathan Will
Jonathan Will and Lauritz Thamsen and Jonathan Bader and Odej Kao
Flora: Efficient Cloud Resource Selection for Big Data Processing via Job Classification
9 pages, 3 figures, 5 tables. Conference: CCGrid 2025
null
null
null
cs.DC
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Distributed dataflow systems like Spark and Flink enable data-parallel processing of large datasets on clusters of cloud resources. Yet, selecting appropriate computational resources for dataflow jobs is often challenging. For efficient execution, individual resource allocations, such as memory and CPU cores, must meet the specific resource demands of the job. Meanwhile, the choices of cloud configurations are often plentiful, especially in public clouds, and the current cost of the available resource options can fluctuate. Addressing this challenge, we present Flora, a low-overhead approach to cost-optimizing cloud cluster configurations for big data processing. Flora lets users categorize jobs according to their data access patterns and derives suitable cluster resource configurations from executions of test jobs of the same category, considering current resource costs. In our evaluation on a new dataset comprising 180 Spark job executions on Google Cloud, Flora's cluster resource selections exhibit an average deviation below 6% from the most cost-optimal solution, with a maximum deviation below 24%.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:40:44 GMT" } ]
2025-03-03T00:00:00
[ [ "Will", "Jonathan", "" ], [ "Thamsen", "Lauritz", "" ], [ "Bader", "Jonathan", "" ], [ "Kao", "Odej", "" ] ]
TITLE: Flora: Efficient Cloud Resource Selection for Big Data Processing via Job Classification ABSTRACT: Distributed dataflow systems like Spark and Flink enable data-parallel processing of large datasets on clusters of cloud resources. Yet, selecting appropriate computational resources for dataflow jobs is often challenging. For efficient execution, individual resource allocations, such as memory and CPU cores, must meet the specific resource demands of the job. Meanwhile, the choices of cloud configurations are often plentiful, especially in public clouds, and the current cost of the available resource options can fluctuate. Addressing this challenge, we present Flora, a low-overhead approach to cost-optimizing cloud cluster configurations for big data processing. Flora lets users categorize jobs according to their data access patterns and derives suitable cluster resource configurations from executions of test jobs of the same category, considering current resource costs. In our evaluation on a new dataset comprising 180 Spark job executions on Google Cloud, Flora's cluster resource selections exhibit an average deviation below 6% from the most cost-optimal solution, with a maximum deviation below 24%.
new_dataset
0.959421
2502.21049
Jingru Fu
Jingru Fu, Yuqi Zheng, Neel Dey, Daniel Ferreira, Rodrigo Moreno
Synthesizing Individualized Aging Brains in Health and Disease with Generative Models and Parallel Transport
20 pages, 9 figures, 6 tables, diffeomorphic registration, parallel transport, brain aging, medical image generation, Alzheimer's disease
null
null
null
cs.CV cs.AI eess.IV
http://creativecommons.org/licenses/by/4.0/
Simulating prospective magnetic resonance imaging (MRI) scans from a given individual brain image is challenging, as it requires accounting for canonical changes in aging and/or disease progression while also considering the individual brain's current status and unique characteristics. While current deep generative models can produce high-resolution anatomically accurate templates for population-wide studies, their ability to predict future aging trajectories for individuals remains limited, particularly in capturing subject-specific neuroanatomical variations over time. In this study, we introduce Individualized Brain Synthesis (InBrainSyn), a framework for synthesizing high-resolution subject-specific longitudinal MRI scans that simulate neurodegeneration in both Alzheimer's disease (AD) and normal aging. InBrainSyn uses a parallel transport algorithm to adapt the population-level aging trajectories learned by a generative deep template network, enabling individualized aging synthesis. As InBrainSyn uses diffeomorphic transformations to simulate aging, the synthesized images are topologically consistent with the original anatomy by design. We evaluated InBrainSyn both quantitatively and qualitatively on AD and healthy control cohorts from the Open Access Series of Imaging Studies - version 3 dataset. Experimentally, InBrainSyn can also model neuroanatomical transitions between normal aging and AD. An evaluation of an external set supports its generalizability. Overall, with only a single baseline scan, InBrainSyn synthesizes realistic 3D spatiotemporal T1w MRI scans, producing personalized longitudinal aging trajectories. The code for InBrainSyn is available at: https://github.com/Fjr9516/InBrainSyn.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:45:09 GMT" } ]
2025-03-03T00:00:00
[ [ "Fu", "Jingru", "" ], [ "Zheng", "Yuqi", "" ], [ "Dey", "Neel", "" ], [ "Ferreira", "Daniel", "" ], [ "Moreno", "Rodrigo", "" ] ]
TITLE: Synthesizing Individualized Aging Brains in Health and Disease with Generative Models and Parallel Transport ABSTRACT: Simulating prospective magnetic resonance imaging (MRI) scans from a given individual brain image is challenging, as it requires accounting for canonical changes in aging and/or disease progression while also considering the individual brain's current status and unique characteristics. While current deep generative models can produce high-resolution anatomically accurate templates for population-wide studies, their ability to predict future aging trajectories for individuals remains limited, particularly in capturing subject-specific neuroanatomical variations over time. In this study, we introduce Individualized Brain Synthesis (InBrainSyn), a framework for synthesizing high-resolution subject-specific longitudinal MRI scans that simulate neurodegeneration in both Alzheimer's disease (AD) and normal aging. InBrainSyn uses a parallel transport algorithm to adapt the population-level aging trajectories learned by a generative deep template network, enabling individualized aging synthesis. As InBrainSyn uses diffeomorphic transformations to simulate aging, the synthesized images are topologically consistent with the original anatomy by design. We evaluated InBrainSyn both quantitatively and qualitatively on AD and healthy control cohorts from the Open Access Series of Imaging Studies - version 3 dataset. Experimentally, InBrainSyn can also model neuroanatomical transitions between normal aging and AD. An evaluation of an external set supports its generalizability. Overall, with only a single baseline scan, InBrainSyn synthesizes realistic 3D spatiotemporal T1w MRI scans, producing personalized longitudinal aging trajectories. The code for InBrainSyn is available at: https://github.com/Fjr9516/InBrainSyn.
no_new_dataset
0.944382
2502.21051
Valentin Guien
Valentin Guien, Violaine Antoine, Romain Lardy, Isabelle Veissier and Luis E C Rocha
Detection of anomalies in cow activity using wavelet transform based features
17 pages, 8 figures, 4 tables, 1 algorithm
null
null
null
cs.LG cs.CE
http://creativecommons.org/licenses/by/4.0/
In Precision Livestock Farming, detecting deviations from optimal or baseline values - i.e. anomalies in time series - is essential to allow undertaking corrective actions rapidly. Here we aim at detecting anomalies in 24h time series of cow activity, with a view to detect cases of disease or oestrus. Deviations must be distinguished from noise which can be very high in case of biological data. It is also important to detect the anomaly early, e.g. before a farmer would notice it visually. Here, we investigate the benefit of using wavelet transforms to denoise data and we assess the performance of an anomaly detection algorithm considering the timing of the detection. We developed features based on the comparisons between the wavelet transforms of the mean of the time series and the wavelet transforms of individual time series instances. We hypothesized that these features contribute to the detection of anomalies in periodic time series using a feature-based algorithm. We tested this hypothesis with two datasets representing cow activity, which typically follows a daily pattern but can deviate due to specific physiological or pathological conditions. We applied features derived from wavelet transform as well as statistical features in an Isolation Forest algorithm. We measured the distance of detection between the days annotated abnormal by animal caretakers days and the days predicted abnormal by the algorithm. The results show that wavelet-based features are among the features most contributing to anomaly detection. They also show that detections are close to the annotated days, and often precede it. In conclusion, using wavelet transforms on time series of cow activity data helps to detect anomalies related to specific cow states. The detection is often obtained on days that precede the day annotated by caretakers, which offer possibility to take corrective actions at an early stage.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:50:18 GMT" } ]
2025-03-03T00:00:00
[ [ "Guien", "Valentin", "" ], [ "Antoine", "Violaine", "" ], [ "Lardy", "Romain", "" ], [ "Veissier", "Isabelle", "" ], [ "Rocha", "Luis E C", "" ] ]
TITLE: Detection of anomalies in cow activity using wavelet transform based features ABSTRACT: In Precision Livestock Farming, detecting deviations from optimal or baseline values - i.e. anomalies in time series - is essential to allow undertaking corrective actions rapidly. Here we aim at detecting anomalies in 24h time series of cow activity, with a view to detect cases of disease or oestrus. Deviations must be distinguished from noise which can be very high in case of biological data. It is also important to detect the anomaly early, e.g. before a farmer would notice it visually. Here, we investigate the benefit of using wavelet transforms to denoise data and we assess the performance of an anomaly detection algorithm considering the timing of the detection. We developed features based on the comparisons between the wavelet transforms of the mean of the time series and the wavelet transforms of individual time series instances. We hypothesized that these features contribute to the detection of anomalies in periodic time series using a feature-based algorithm. We tested this hypothesis with two datasets representing cow activity, which typically follows a daily pattern but can deviate due to specific physiological or pathological conditions. We applied features derived from wavelet transform as well as statistical features in an Isolation Forest algorithm. We measured the distance of detection between the days annotated abnormal by animal caretakers days and the days predicted abnormal by the algorithm. The results show that wavelet-based features are among the features most contributing to anomaly detection. They also show that detections are close to the annotated days, and often precede it. In conclusion, using wavelet transforms on time series of cow activity data helps to detect anomalies related to specific cow states. The detection is often obtained on days that precede the day annotated by caretakers, which offer possibility to take corrective actions at an early stage.
no_new_dataset
0.94801
2502.21054
Emanuele Vivoli
Emanuele Vivoli, Lorenzo Capineri, Marco Bertini
HoloMine: A Synthetic Dataset for Buried Landmines Recognition using Microwave Holographic Imaging
under review
null
null
null
cs.CV eess.IV eess.SP
http://creativecommons.org/licenses/by/4.0/
The detection and removal of landmines is a complex and risky task that requires advanced remote sensing techniques to reduce the risk for the professionals involved in this task. In this paper, we propose a novel synthetic dataset for buried landmine detection to provide researchers with a valuable resource to observe, measure, locate, and address issues in landmine detection. The dataset consists of 41,800 microwave holographic images (2D) and their holographic inverted scans (3D) of different types of buried objects, including landmines, clutter, and pottery objects, and is collected by means of a microwave holography sensor. We evaluate the performance of several state-of-the-art deep learning models trained on our synthetic dataset for various classification tasks. While the results do not yield yet high performances, showing the difficulty of the proposed task, we believe that our dataset has significant potential to drive progress in the field of landmine detection thanks to the accuracy and resolution obtainable using holographic radars. To the best of our knowledge, our dataset is the first of its kind and will help drive further research on computer vision methods to automatize mine detection, with the overall goal of reducing the risks and the costs of the demining process.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:53:35 GMT" } ]
2025-03-03T00:00:00
[ [ "Vivoli", "Emanuele", "" ], [ "Capineri", "Lorenzo", "" ], [ "Bertini", "Marco", "" ] ]
TITLE: HoloMine: A Synthetic Dataset for Buried Landmines Recognition using Microwave Holographic Imaging ABSTRACT: The detection and removal of landmines is a complex and risky task that requires advanced remote sensing techniques to reduce the risk for the professionals involved in this task. In this paper, we propose a novel synthetic dataset for buried landmine detection to provide researchers with a valuable resource to observe, measure, locate, and address issues in landmine detection. The dataset consists of 41,800 microwave holographic images (2D) and their holographic inverted scans (3D) of different types of buried objects, including landmines, clutter, and pottery objects, and is collected by means of a microwave holography sensor. We evaluate the performance of several state-of-the-art deep learning models trained on our synthetic dataset for various classification tasks. While the results do not yield yet high performances, showing the difficulty of the proposed task, we believe that our dataset has significant potential to drive progress in the field of landmine detection thanks to the accuracy and resolution obtainable using holographic radars. To the best of our knowledge, our dataset is the first of its kind and will help drive further research on computer vision methods to automatize mine detection, with the overall goal of reducing the risks and the costs of the demining process.
new_dataset
0.961642
2502.21055
Micha{\l} Romaszewski
Przemys{\l}aw Seku{\l}a, Micha{\l} Romaszewski, Przemys{\l}aw G{\l}omb, Micha{\l} Cholewa, {\L}ukasz Pawela
Quantum-aware Transformer model for state classification
13 pages, 1 figure
null
null
null
quant-ph cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Entanglement is a fundamental feature of quantum mechanics, playing a crucial role in quantum information processing. However, classifying entangled states, particularly in the mixed-state regime, remains a challenging problem, especially as system dimensions increase. In this work, we focus on bipartite quantum states and present a data-driven approach to entanglement classification using transformer-based neural networks. Our dataset consists of a diverse set of bipartite states, including pure separable states, Werner entangled states, general entangled states, and maximally entangled states. We pretrain the transformer in an unsupervised fashion by masking elements of vectorized Hermitian matrix representations of quantum states, allowing the model to learn structural properties of quantum density matrices. This approach enables the model to generalize entanglement characteristics across different classes of states. Once trained, our method achieves near-perfect classification accuracy, effectively distinguishing between separable and entangled states. Compared to previous Machine Learning, our method successfully adapts transformers for quantum state analysis, demonstrating their ability to systematically identify entanglement in bipartite systems. These results highlight the potential of modern machine learning techniques in automating entanglement detection and classification, bridging the gap between quantum information theory and artificial intelligence.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:56:48 GMT" } ]
2025-03-03T00:00:00
[ [ "Sekuła", "Przemysław", "" ], [ "Romaszewski", "Michał", "" ], [ "Głomb", "Przemysław", "" ], [ "Cholewa", "Michał", "" ], [ "Pawela", "Łukasz", "" ] ]
TITLE: Quantum-aware Transformer model for state classification ABSTRACT: Entanglement is a fundamental feature of quantum mechanics, playing a crucial role in quantum information processing. However, classifying entangled states, particularly in the mixed-state regime, remains a challenging problem, especially as system dimensions increase. In this work, we focus on bipartite quantum states and present a data-driven approach to entanglement classification using transformer-based neural networks. Our dataset consists of a diverse set of bipartite states, including pure separable states, Werner entangled states, general entangled states, and maximally entangled states. We pretrain the transformer in an unsupervised fashion by masking elements of vectorized Hermitian matrix representations of quantum states, allowing the model to learn structural properties of quantum density matrices. This approach enables the model to generalize entanglement characteristics across different classes of states. Once trained, our method achieves near-perfect classification accuracy, effectively distinguishing between separable and entangled states. Compared to previous Machine Learning, our method successfully adapts transformers for quantum state analysis, demonstrating their ability to systematically identify entanglement in bipartite systems. These results highlight the potential of modern machine learning techniques in automating entanglement detection and classification, bridging the gap between quantum information theory and artificial intelligence.
new_dataset
0.908374
2502.21059
Zhen Sun
Ziyi Zhang, Zhen Sun, Zongmin Zhang, Jihui Guo, Xinlei He
FC-Attack: Jailbreaking Large Vision-Language Models via Auto-Generated Flowcharts
13 pages, 6 figures
null
null
null
cs.CV cs.AI cs.CR cs.LG
http://creativecommons.org/licenses/by/4.0/
Large Vision-Language Models (LVLMs) have become powerful and widely adopted in some practical applications. However, recent research has revealed their vulnerability to multimodal jailbreak attacks, whereby the model can be induced to generate harmful content, leading to safety risks. Although most LVLMs have undergone safety alignment, recent research shows that the visual modality is still vulnerable to jailbreak attacks. In our work, we discover that by using flowcharts with partially harmful information, LVLMs can be induced to provide additional harmful details. Based on this, we propose a jailbreak attack method based on auto-generated flowcharts, FC-Attack. Specifically, FC-Attack first fine-tunes a pre-trained LLM to create a step-description generator based on benign datasets. The generator is then used to produce step descriptions corresponding to a harmful query, which are transformed into flowcharts in 3 different shapes (vertical, horizontal, and S-shaped) as visual prompts. These flowcharts are then combined with a benign textual prompt to execute a jailbreak attack on LVLMs. Our evaluations using the Advbench dataset show that FC-Attack achieves over 90% attack success rates on Gemini-1.5, Llaval-Next, Qwen2-VL, and InternVL-2.5 models, outperforming existing LVLM jailbreak methods. Additionally, we investigate factors affecting the attack performance, including the number of steps and the font styles in the flowcharts. Our evaluation shows that FC-Attack can improve the jailbreak performance from 4% to 28% in Claude-3.5 by changing the font style. To mitigate the attack, we explore several defenses and find that AdaShield can largely reduce the jailbreak performance but with the cost of utility drop.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 13:59:11 GMT" } ]
2025-03-03T00:00:00
[ [ "Zhang", "Ziyi", "" ], [ "Sun", "Zhen", "" ], [ "Zhang", "Zongmin", "" ], [ "Guo", "Jihui", "" ], [ "He", "Xinlei", "" ] ]
TITLE: FC-Attack: Jailbreaking Large Vision-Language Models via Auto-Generated Flowcharts ABSTRACT: Large Vision-Language Models (LVLMs) have become powerful and widely adopted in some practical applications. However, recent research has revealed their vulnerability to multimodal jailbreak attacks, whereby the model can be induced to generate harmful content, leading to safety risks. Although most LVLMs have undergone safety alignment, recent research shows that the visual modality is still vulnerable to jailbreak attacks. In our work, we discover that by using flowcharts with partially harmful information, LVLMs can be induced to provide additional harmful details. Based on this, we propose a jailbreak attack method based on auto-generated flowcharts, FC-Attack. Specifically, FC-Attack first fine-tunes a pre-trained LLM to create a step-description generator based on benign datasets. The generator is then used to produce step descriptions corresponding to a harmful query, which are transformed into flowcharts in 3 different shapes (vertical, horizontal, and S-shaped) as visual prompts. These flowcharts are then combined with a benign textual prompt to execute a jailbreak attack on LVLMs. Our evaluations using the Advbench dataset show that FC-Attack achieves over 90% attack success rates on Gemini-1.5, Llaval-Next, Qwen2-VL, and InternVL-2.5 models, outperforming existing LVLM jailbreak methods. Additionally, we investigate factors affecting the attack performance, including the number of steps and the font styles in the flowcharts. Our evaluation shows that FC-Attack can improve the jailbreak performance from 4% to 28% in Claude-3.5 by changing the font style. To mitigate the attack, we explore several defenses and find that AdaShield can largely reduce the jailbreak performance but with the cost of utility drop.
no_new_dataset
0.923868
2502.21074
Zhenyi Shen
Zhenyi Shen, Hanqi Yan, Linhai Zhang, Zhanghao Hu, Yali Du, Yulan He
CODI: Compressing Chain-of-Thought into Continuous Space via Self-Distillation
15 pages
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
Chain-of-Thought (CoT) enhances Large Language Models (LLMs) by enabling step-by-step reasoning in natural language. However, the language space may be suboptimal for reasoning. While implicit CoT methods attempt to enable reasoning without explicit CoT tokens, they have consistently lagged behind explicit CoT method in task performance. We propose CODI (Continuous Chain-of-Thought via Self-Distillation), a novel framework that distills CoT into a continuous space, where a shared model acts as both teacher and student, jointly learning explicit and implicit CoT while aligning their hidden activation on the token generating the final answer. CODI is the first implicit CoT method to match explicit CoT's performance on GSM8k while achieving 3.1x compression, surpassing the previous state-of-the-art by 28.2% in accuracy. Furthermore, CODI demonstrates scalability, robustness, and generalizability to more complex CoT datasets. Additionally, CODI retains interpretability by decoding its continuous thoughts, making its reasoning process transparent. Our findings establish implicit CoT as not only a more efficient but a powerful alternative to explicit CoT.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:07:48 GMT" } ]
2025-03-03T00:00:00
[ [ "Shen", "Zhenyi", "" ], [ "Yan", "Hanqi", "" ], [ "Zhang", "Linhai", "" ], [ "Hu", "Zhanghao", "" ], [ "Du", "Yali", "" ], [ "He", "Yulan", "" ] ]
TITLE: CODI: Compressing Chain-of-Thought into Continuous Space via Self-Distillation ABSTRACT: Chain-of-Thought (CoT) enhances Large Language Models (LLMs) by enabling step-by-step reasoning in natural language. However, the language space may be suboptimal for reasoning. While implicit CoT methods attempt to enable reasoning without explicit CoT tokens, they have consistently lagged behind explicit CoT method in task performance. We propose CODI (Continuous Chain-of-Thought via Self-Distillation), a novel framework that distills CoT into a continuous space, where a shared model acts as both teacher and student, jointly learning explicit and implicit CoT while aligning their hidden activation on the token generating the final answer. CODI is the first implicit CoT method to match explicit CoT's performance on GSM8k while achieving 3.1x compression, surpassing the previous state-of-the-art by 28.2% in accuracy. Furthermore, CODI demonstrates scalability, robustness, and generalizability to more complex CoT datasets. Additionally, CODI retains interpretability by decoding its continuous thoughts, making its reasoning process transparent. Our findings establish implicit CoT as not only a more efficient but a powerful alternative to explicit CoT.
no_new_dataset
0.945751
2502.21079
Yifei Xia
Yifei Xia, Suhan Ling, Fangcheng Fu, Yujie Wang, Huixia Li, Xuefeng Xiao, Bin Cui
Training-free and Adaptive Sparse Attention for Efficient Long Video Generation
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
Generating high-fidelity long videos with Diffusion Transformers (DiTs) is often hindered by significant latency, primarily due to the computational demands of attention mechanisms. For instance, generating an 8-second 720p video (110K tokens) with HunyuanVideo takes about 600 PFLOPs, with around 500 PFLOPs consumed by attention computations. To address this issue, we propose AdaSpa, the first Dynamic Pattern and Online Precise Search sparse attention method. Firstly, to realize the Dynamic Pattern, we introduce a blockified pattern to efficiently capture the hierarchical sparsity inherent in DiTs. This is based on our observation that sparse characteristics of DiTs exhibit hierarchical and blockified structures between and within different modalities. This blockified approach significantly reduces the complexity of attention computation while maintaining high fidelity in the generated videos. Secondly, to enable Online Precise Search, we propose the Fused LSE-Cached Search with Head-adaptive Hierarchical Block Sparse Attention. This method is motivated by our finding that DiTs' sparse pattern and LSE vary w.r.t. inputs, layers, and heads, but remain invariant across denoising steps. By leveraging this invariance across denoising steps, it adapts to the dynamic nature of DiTs and allows for precise, real-time identification of sparse indices with minimal overhead. AdaSpa is implemented as an adaptive, plug-and-play solution and can be integrated seamlessly with existing DiTs, requiring neither additional fine-tuning nor a dataset-dependent profiling. Extensive experiments validate that AdaSpa delivers substantial acceleration across various models while preserving video quality, establishing itself as a robust and scalable approach to efficient video generation.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:11:20 GMT" } ]
2025-03-03T00:00:00
[ [ "Xia", "Yifei", "" ], [ "Ling", "Suhan", "" ], [ "Fu", "Fangcheng", "" ], [ "Wang", "Yujie", "" ], [ "Li", "Huixia", "" ], [ "Xiao", "Xuefeng", "" ], [ "Cui", "Bin", "" ] ]
TITLE: Training-free and Adaptive Sparse Attention for Efficient Long Video Generation ABSTRACT: Generating high-fidelity long videos with Diffusion Transformers (DiTs) is often hindered by significant latency, primarily due to the computational demands of attention mechanisms. For instance, generating an 8-second 720p video (110K tokens) with HunyuanVideo takes about 600 PFLOPs, with around 500 PFLOPs consumed by attention computations. To address this issue, we propose AdaSpa, the first Dynamic Pattern and Online Precise Search sparse attention method. Firstly, to realize the Dynamic Pattern, we introduce a blockified pattern to efficiently capture the hierarchical sparsity inherent in DiTs. This is based on our observation that sparse characteristics of DiTs exhibit hierarchical and blockified structures between and within different modalities. This blockified approach significantly reduces the complexity of attention computation while maintaining high fidelity in the generated videos. Secondly, to enable Online Precise Search, we propose the Fused LSE-Cached Search with Head-adaptive Hierarchical Block Sparse Attention. This method is motivated by our finding that DiTs' sparse pattern and LSE vary w.r.t. inputs, layers, and heads, but remain invariant across denoising steps. By leveraging this invariance across denoising steps, it adapts to the dynamic nature of DiTs and allows for precise, real-time identification of sparse indices with minimal overhead. AdaSpa is implemented as an adaptive, plug-and-play solution and can be integrated seamlessly with existing DiTs, requiring neither additional fine-tuning nor a dataset-dependent profiling. Extensive experiments validate that AdaSpa delivers substantial acceleration across various models while preserving video quality, establishing itself as a robust and scalable approach to efficient video generation.
no_new_dataset
0.946745
2502.21085
Jing-Yuan Chang
Jing-Yuan Chang
BST: Badminton Stroke-type Transformer for Skeleton-based Action Recognition in Racket Sports
8 pages (excluding references). The code will be released in a few months
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Badminton, known for having the fastest ball speeds among all sports, presents significant challenges to the field of computer vision, including player identification, court line detection, shuttlecock trajectory tracking, and player stroke-type classification. In this paper, we introduce a novel video segmentation strategy to extract frames of each player's racket swing in a badminton broadcast match. These segmented frames are then processed by two existing models: one for Human Pose Estimation to obtain player skeletal joints, and the other for shuttlecock trajectory detection to extract shuttlecock trajectories. Leveraging these joints, trajectories, and player positions as inputs, we propose Badminton Stroke-type Transformer (BST) to classify player stroke-types in singles. To the best of our knowledge, experimental results demonstrate that our method outperforms the previous state-of-the-art on the largest publicly available badminton video dataset, ShuttleSet, which shows that effectively leveraging ball trajectory is likely to be a trend for racket sports action recognition.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:18:39 GMT" } ]
2025-03-03T00:00:00
[ [ "Chang", "Jing-Yuan", "" ] ]
TITLE: BST: Badminton Stroke-type Transformer for Skeleton-based Action Recognition in Racket Sports ABSTRACT: Badminton, known for having the fastest ball speeds among all sports, presents significant challenges to the field of computer vision, including player identification, court line detection, shuttlecock trajectory tracking, and player stroke-type classification. In this paper, we introduce a novel video segmentation strategy to extract frames of each player's racket swing in a badminton broadcast match. These segmented frames are then processed by two existing models: one for Human Pose Estimation to obtain player skeletal joints, and the other for shuttlecock trajectory detection to extract shuttlecock trajectories. Leveraging these joints, trajectories, and player positions as inputs, we propose Badminton Stroke-type Transformer (BST) to classify player stroke-types in singles. To the best of our knowledge, experimental results demonstrate that our method outperforms the previous state-of-the-art on the largest publicly available badminton video dataset, ShuttleSet, which shows that effectively leveraging ball trajectory is likely to be a trend for racket sports action recognition.
no_new_dataset
0.951006
2502.21086
\"Ozg\"un Turgut
\"Ozg\"un Turgut, Felix S. Bott, Markus Ploner, Daniel Rueckert
Are foundation models useful feature extractors for electroencephalography analysis?
null
null
null
null
cs.AI cs.LG
http://creativecommons.org/licenses/by/4.0/
The success of foundation models in natural language processing and computer vision has motivated similar approaches for general time series analysis. While these models are effective for a variety of tasks, their applicability in medical domains with limited data remains largely unexplored. To address this, we investigate the effectiveness of foundation models in medical time series analysis involving electroencephalography (EEG). Through extensive experiments on tasks such as age prediction, seizure detection, and the classification of clinically relevant EEG events, we compare their diagnostic accuracy with that of specialised EEG models. Our analysis shows that foundation models extract meaningful EEG features, outperform specialised models even without domain adaptation, and localise task-specific biomarkers. Moreover, we demonstrate that diagnostic accuracy is substantially influenced by architectural choices such as context length. Overall, our study reveals that foundation models with general time series understanding eliminate the dependency on large domain-specific datasets, making them valuable tools for clinical practice.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:21:34 GMT" } ]
2025-03-03T00:00:00
[ [ "Turgut", "Özgün", "" ], [ "Bott", "Felix S.", "" ], [ "Ploner", "Markus", "" ], [ "Rueckert", "Daniel", "" ] ]
TITLE: Are foundation models useful feature extractors for electroencephalography analysis? ABSTRACT: The success of foundation models in natural language processing and computer vision has motivated similar approaches for general time series analysis. While these models are effective for a variety of tasks, their applicability in medical domains with limited data remains largely unexplored. To address this, we investigate the effectiveness of foundation models in medical time series analysis involving electroencephalography (EEG). Through extensive experiments on tasks such as age prediction, seizure detection, and the classification of clinically relevant EEG events, we compare their diagnostic accuracy with that of specialised EEG models. Our analysis shows that foundation models extract meaningful EEG features, outperform specialised models even without domain adaptation, and localise task-specific biomarkers. Moreover, we demonstrate that diagnostic accuracy is substantially influenced by architectural choices such as context length. Overall, our study reveals that foundation models with general time series understanding eliminate the dependency on large domain-specific datasets, making them valuable tools for clinical practice.
no_new_dataset
0.948917
2502.21087
Hansi Yang
Hansi Yang, Qi Zhang, Wei Jiang, Jianguo Li
PASemiQA: Plan-Assisted Agent for Question Answering on Semi-Structured Data with Text and Relational Information
null
null
null
null
cs.CL cs.AI
http://creativecommons.org/licenses/by-nc-sa/4.0/
Large language models (LLMs) have shown impressive abilities in answering questions across various domains, but they often encounter hallucination issues on questions that require professional and up-to-date knowledge. To address this limitation, retrieval-augmented generation (RAG) techniques have been proposed, which retrieve relevant information from external sources to inform their responses. However, existing RAG methods typically focus on a single type of external data, such as vectorized text database or knowledge graphs, and cannot well handle real-world questions on semi-structured data containing both text and relational information. To bridge this gap, we introduce PASemiQA, a novel approach that jointly leverages text and relational information in semi-structured data to answer questions. PASemiQA first generates a plan to identify relevant text and relational information to answer the question in semi-structured data, and then uses an LLM agent to traverse the semi-structured data and extract necessary information. Our empirical results demonstrate the effectiveness of PASemiQA across different semi-structured datasets from various domains, showcasing its potential to improve the accuracy and reliability of question answering systems on semi-structured data.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:26:47 GMT" } ]
2025-03-03T00:00:00
[ [ "Yang", "Hansi", "" ], [ "Zhang", "Qi", "" ], [ "Jiang", "Wei", "" ], [ "Li", "Jianguo", "" ] ]
TITLE: PASemiQA: Plan-Assisted Agent for Question Answering on Semi-Structured Data with Text and Relational Information ABSTRACT: Large language models (LLMs) have shown impressive abilities in answering questions across various domains, but they often encounter hallucination issues on questions that require professional and up-to-date knowledge. To address this limitation, retrieval-augmented generation (RAG) techniques have been proposed, which retrieve relevant information from external sources to inform their responses. However, existing RAG methods typically focus on a single type of external data, such as vectorized text database or knowledge graphs, and cannot well handle real-world questions on semi-structured data containing both text and relational information. To bridge this gap, we introduce PASemiQA, a novel approach that jointly leverages text and relational information in semi-structured data to answer questions. PASemiQA first generates a plan to identify relevant text and relational information to answer the question in semi-structured data, and then uses an LLM agent to traverse the semi-structured data and extract necessary information. Our empirical results demonstrate the effectiveness of PASemiQA across different semi-structured datasets from various domains, showcasing its potential to improve the accuracy and reliability of question answering systems on semi-structured data.
no_new_dataset
0.94474
2502.21110
Charles Dawson
Charles Dawson, Van Tran, Max Z. Li, Chuchu Fan
Rare event modeling with self-regularized normalizing flows: what can we learn from a single failure?
Published at ICLR 2025
null
null
null
cs.LG stat.ML
http://creativecommons.org/licenses/by-sa/4.0/
Increased deployment of autonomous systems in fields like transportation and robotics have seen a corresponding increase in safety-critical failures. These failures can be difficult to model and debug due to the relative lack of data: compared to tens of thousands of examples from normal operations, we may have only seconds of data leading up to the failure. This scarcity makes it challenging to train generative models of rare failure events, as existing methods risk either overfitting to noise in the limited failure dataset or underfitting due to an overly strong prior. We address this challenge with CalNF, or calibrated normalizing flows, a self-regularized framework for posterior learning from limited data. CalNF achieves state-of-the-art performance on data-limited failure modeling and inverse problems and enables a first-of-a-kind case study into the root causes of the 2022 Southwest Airlines scheduling crisis.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:47:52 GMT" } ]
2025-03-03T00:00:00
[ [ "Dawson", "Charles", "" ], [ "Tran", "Van", "" ], [ "Li", "Max Z.", "" ], [ "Fan", "Chuchu", "" ] ]
TITLE: Rare event modeling with self-regularized normalizing flows: what can we learn from a single failure? ABSTRACT: Increased deployment of autonomous systems in fields like transportation and robotics have seen a corresponding increase in safety-critical failures. These failures can be difficult to model and debug due to the relative lack of data: compared to tens of thousands of examples from normal operations, we may have only seconds of data leading up to the failure. This scarcity makes it challenging to train generative models of rare failure events, as existing methods risk either overfitting to noise in the limited failure dataset or underfitting due to an overly strong prior. We address this challenge with CalNF, or calibrated normalizing flows, a self-regularized framework for posterior learning from limited data. CalNF achieves state-of-the-art performance on data-limited failure modeling and inverse problems and enables a first-of-a-kind case study into the root causes of the 2022 Southwest Airlines scheduling crisis.
no_new_dataset
0.949342
2502.21112
Francesco Osborne
Mattia Birti, Francesco Osborne, Andrea Maurino
Optimizing Large Language Models for ESG Activity Detection in Financial Texts
null
null
null
null
cs.AI cs.CE cs.CL cs.CY cs.IR
http://creativecommons.org/licenses/by/4.0/
The integration of Environmental, Social, and Governance (ESG) factors into corporate decision-making is a fundamental aspect of sustainable finance. However, ensuring that business practices align with evolving regulatory frameworks remains a persistent challenge. AI-driven solutions for automatically assessing the alignment of sustainability reports and non-financial disclosures with specific ESG activities could greatly support this process. Yet, this task remains complex due to the limitations of general-purpose Large Language Models (LLMs) in domain-specific contexts and the scarcity of structured, high-quality datasets. In this paper, we investigate the ability of current-generation LLMs to identify text related to environmental activities. Furthermore, we demonstrate that their performance can be significantly enhanced through fine-tuning on a combination of original and synthetically generated data. To this end, we introduce ESG-Activities, a benchmark dataset containing 1,325 labelled text segments classified according to the EU ESG taxonomy. Our experimental results show that fine-tuning on ESG-Activities significantly enhances classification accuracy, with open models such as Llama 7B and Gemma 7B outperforming large proprietary solutions in specific configurations. These findings have important implications for financial analysts, policymakers, and AI researchers seeking to enhance ESG transparency and compliance through advanced natural language processing techniques.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:52:25 GMT" } ]
2025-03-03T00:00:00
[ [ "Birti", "Mattia", "" ], [ "Osborne", "Francesco", "" ], [ "Maurino", "Andrea", "" ] ]
TITLE: Optimizing Large Language Models for ESG Activity Detection in Financial Texts ABSTRACT: The integration of Environmental, Social, and Governance (ESG) factors into corporate decision-making is a fundamental aspect of sustainable finance. However, ensuring that business practices align with evolving regulatory frameworks remains a persistent challenge. AI-driven solutions for automatically assessing the alignment of sustainability reports and non-financial disclosures with specific ESG activities could greatly support this process. Yet, this task remains complex due to the limitations of general-purpose Large Language Models (LLMs) in domain-specific contexts and the scarcity of structured, high-quality datasets. In this paper, we investigate the ability of current-generation LLMs to identify text related to environmental activities. Furthermore, we demonstrate that their performance can be significantly enhanced through fine-tuning on a combination of original and synthetically generated data. To this end, we introduce ESG-Activities, a benchmark dataset containing 1,325 labelled text segments classified according to the EU ESG taxonomy. Our experimental results show that fine-tuning on ESG-Activities significantly enhances classification accuracy, with open models such as Llama 7B and Gemma 7B outperforming large proprietary solutions in specific configurations. These findings have important implications for financial analysts, policymakers, and AI researchers seeking to enhance ESG transparency and compliance through advanced natural language processing techniques.
new_dataset
0.963643
2502.21120
Yunfan Lu
Yunfan Lu, Xiaogang Xu, Hao Lu, Yanlin Qian, Pengteng Li, Huizai Yao, Bin Yang, Junyi Li, Qianyi Cai, Weiyu Guo, Hui Xiong
SEE: See Everything Every Time -- Adaptive Brightness Adjustment for Broad Light Range Images via Events
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Event cameras, with a high dynamic range exceeding $120dB$, significantly outperform traditional embedded cameras, robustly recording detailed changing information under various lighting conditions, including both low- and high-light situations. However, recent research on utilizing event data has primarily focused on low-light image enhancement, neglecting image enhancement and brightness adjustment across a broader range of lighting conditions, such as normal or high illumination. Based on this, we propose a novel research question: how to employ events to enhance and adaptively adjust the brightness of images captured under broad lighting conditions? To investigate this question, we first collected a new dataset, SEE-600K, consisting of 610,126 images and corresponding events across 202 scenarios, each featuring an average of four lighting conditions with over a 1000-fold variation in illumination. Subsequently, we propose a framework that effectively utilizes events to smoothly adjust image brightness through the use of prompts. Our framework captures color through sensor patterns, uses cross-attention to model events as a brightness dictionary, and adjusts the image's dynamic range to form a broad light-range representation (BLR), which is then decoded at the pixel level based on the brightness prompt. Experimental results demonstrate that our method not only performs well on the low-light enhancement dataset but also shows robust performance on broader light-range image enhancement using the SEE-600K dataset. Additionally, our approach enables pixel-level brightness adjustment, providing flexibility for post-processing and inspiring more imaging applications. The dataset and source code are publicly available at:https://github.com/yunfanLu/SEE.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 14:55:37 GMT" } ]
2025-03-03T00:00:00
[ [ "Lu", "Yunfan", "" ], [ "Xu", "Xiaogang", "" ], [ "Lu", "Hao", "" ], [ "Qian", "Yanlin", "" ], [ "Li", "Pengteng", "" ], [ "Yao", "Huizai", "" ], [ "Yang", "Bin", "" ], [ "Li", "Junyi", "" ], [ "Cai", "Qianyi", "" ], [ "Guo", "Weiyu", "" ], [ "Xiong", "Hui", "" ] ]
TITLE: SEE: See Everything Every Time -- Adaptive Brightness Adjustment for Broad Light Range Images via Events ABSTRACT: Event cameras, with a high dynamic range exceeding $120dB$, significantly outperform traditional embedded cameras, robustly recording detailed changing information under various lighting conditions, including both low- and high-light situations. However, recent research on utilizing event data has primarily focused on low-light image enhancement, neglecting image enhancement and brightness adjustment across a broader range of lighting conditions, such as normal or high illumination. Based on this, we propose a novel research question: how to employ events to enhance and adaptively adjust the brightness of images captured under broad lighting conditions? To investigate this question, we first collected a new dataset, SEE-600K, consisting of 610,126 images and corresponding events across 202 scenarios, each featuring an average of four lighting conditions with over a 1000-fold variation in illumination. Subsequently, we propose a framework that effectively utilizes events to smoothly adjust image brightness through the use of prompts. Our framework captures color through sensor patterns, uses cross-attention to model events as a brightness dictionary, and adjusts the image's dynamic range to form a broad light-range representation (BLR), which is then decoded at the pixel level based on the brightness prompt. Experimental results demonstrate that our method not only performs well on the low-light enhancement dataset but also shows robust performance on broader light-range image enhancement using the SEE-600K dataset. Additionally, our approach enables pixel-level brightness adjustment, providing flexibility for post-processing and inspiring more imaging applications. The dataset and source code are publicly available at:https://github.com/yunfanLu/SEE.
new_dataset
0.957991
2502.21143
Hyungi Lee
Hyungi Lee, Seungyoo Lee, Juho Lee
Variational Bayesian Pseudo-Coreset
The Thirteenth International Conference on Learning Representations (ICLR2025)
null
null
null
cs.LG
http://creativecommons.org/licenses/by/4.0/
The success of deep learning requires large datasets and extensive training, which can create significant computational challenges. To address these challenges, pseudo-coresets, small learnable datasets that mimic the entire data, have been proposed. Bayesian Neural Networks, which offer predictive uncertainty and probabilistic interpretation for deep neural networks, also face issues with large-scale datasets due to their high-dimensional parameter space. Prior works on Bayesian Pseudo-Coresets (BPC) attempt to reduce the computational load for computing weight posterior distribution by a small number of pseudo-coresets but suffer from memory inefficiency during BPC training and sub-optimal results. To overcome these limitations, we propose Variational Bayesian Pseudo-Coreset (VBPC), a novel approach that utilizes variational inference to efficiently approximate the posterior distribution, reducing memory usage and computational costs while improving performance across benchmark datasets.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 15:26:10 GMT" } ]
2025-03-03T00:00:00
[ [ "Lee", "Hyungi", "" ], [ "Lee", "Seungyoo", "" ], [ "Lee", "Juho", "" ] ]
TITLE: Variational Bayesian Pseudo-Coreset ABSTRACT: The success of deep learning requires large datasets and extensive training, which can create significant computational challenges. To address these challenges, pseudo-coresets, small learnable datasets that mimic the entire data, have been proposed. Bayesian Neural Networks, which offer predictive uncertainty and probabilistic interpretation for deep neural networks, also face issues with large-scale datasets due to their high-dimensional parameter space. Prior works on Bayesian Pseudo-Coresets (BPC) attempt to reduce the computational load for computing weight posterior distribution by a small number of pseudo-coresets but suffer from memory inefficiency during BPC training and sub-optimal results. To overcome these limitations, we propose Variational Bayesian Pseudo-Coreset (VBPC), a novel approach that utilizes variational inference to efficiently approximate the posterior distribution, reducing memory usage and computational costs while improving performance across benchmark datasets.
no_new_dataset
0.952175
2502.21147
Eli Verwimp
Eli Verwimp, Guy Hacohen and Tinne Tuytelaars
Same accuracy, twice as fast: continuous training surpasses retraining from scratch
null
null
null
null
cs.LG cs.CV
http://creativecommons.org/licenses/by/4.0/
Continual learning aims to enable models to adapt to new datasets without losing performance on previously learned data, often assuming that prior data is no longer available. However, in many practical scenarios, both old and new data are accessible. In such cases, good performance on both datasets is typically achieved by abandoning the model trained on the previous data and re-training a new model from scratch on both datasets. This training from scratch is computationally expensive. In contrast, methods that leverage the previously trained model and old data are worthy of investigation, as they could significantly reduce computational costs. Our evaluation framework quantifies the computational savings of such methods while maintaining or exceeding the performance of training from scratch. We identify key optimization aspects -- initialization, regularization, data selection, and hyper-parameters -- that can each contribute to reducing computational costs. For each aspect, we propose effective first-step methods that already yield substantial computational savings. By combining these methods, we achieve up to 2.7x reductions in computation time across various computer vision tasks, highlighting the potential for further advancements in this area.
[ { "version": "v1", "created": "Fri, 28 Feb 2025 15:28:12 GMT" } ]
2025-03-03T00:00:00
[ [ "Verwimp", "Eli", "" ], [ "Hacohen", "Guy", "" ], [ "Tuytelaars", "Tinne", "" ] ]
TITLE: Same accuracy, twice as fast: continuous training surpasses retraining from scratch ABSTRACT: Continual learning aims to enable models to adapt to new datasets without losing performance on previously learned data, often assuming that prior data is no longer available. However, in many practical scenarios, both old and new data are accessible. In such cases, good performance on both datasets is typically achieved by abandoning the model trained on the previous data and re-training a new model from scratch on both datasets. This training from scratch is computationally expensive. In contrast, methods that leverage the previously trained model and old data are worthy of investigation, as they could significantly reduce computational costs. Our evaluation framework quantifies the computational savings of such methods while maintaining or exceeding the performance of training from scratch. We identify key optimization aspects -- initialization, regularization, data selection, and hyper-parameters -- that can each contribute to reducing computational costs. For each aspect, we propose effective first-step methods that already yield substantial computational savings. By combining these methods, we achieve up to 2.7x reductions in computation time across various computer vision tasks, highlighting the potential for further advancements in this area.
no_new_dataset
0.945601