Yapeng Tian
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yapeng Tian has a total of 170 co-authors. Here are the top co-authors by number of shared publications:
- Chenliang Xu
Department of Computer Science, University of Rochester, Rochester, USA
14 shared publications - Wenming Yang
Tsinghua University
11 shared publications - Qingmin Liao
Shenzhen International Graduate School, Tsinghua University, Shenzhen, China
7 shared publications - Yulun Zhang
School of Computer Science, Shanghai Jiao Tong University, Shanghai, China
7 shared publications - Shijian Deng
The University of Texas at Dallas
6 shared publications - Yuhang Zhao
Department of Computer Sciences, University of Wisconsin-Madison, Madison, USA
6 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1664-1673
Talking to Me or Someone Else? Rethinking Talk-to-Me Detection in Egocentric Videos
Online understanding of who is talking to the camera wearer is a key capability for egocentric social interaction. However, existing talk-to-me (TTM) studies are commonly formulated as offline clip-level recognition, which is poorly aligned with online interaction and overlooks the diverse non-TTM speaking states that naturally arise in egocentric videos. In this paper, we revisit this problem by reformulating it as an online, frame-level prediction task. Instead of treating TTM as a binary problem against a single negative class, we model it in the presence of diverse and previously underexplored non-TTM states, such as talking-to-others, self-talking, and background conditions. To support this new formulation, we construct an Online TTM Dataset consisting of 406 egocentric video clips with approximately 900K annotated frames, each labeled with frame-level social interaction categories (e.g., background, TTM, talking-to-others, self-talking), by extending the Ego4D social interaction benchmark. In this benchmark, we evaluate five adapted baselines and develop a new model that integrates social cues across modalities. Experimental results show that our multimodal model, which jointly leverages audio, visual, and speech-semantic cues, achieves 75.5% frame-level F1 on TTM, outperforming strong baselines and enabling a systematic analysis of how different speaking states affect TTM recognition.
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1664-1673
Open AccessTalking to Me or Someone Else? Rethinking Talk-to-Me Detection in Egocentric Videos
Online understanding of who is talking to the camera wearer is a key capability for egocentric social interaction. However, existing talk-to-me (TTM) studies are commonly formulated as offline clip-level recognition, which is poorly aligned with online interaction and overlooks the diverse non-TTM speaking states that naturally arise in egocentric videos. In this paper, we revisit this problem by reformulating it as an online, frame-level prediction task. Instead of treating TTM as a binary problem against a single negative class, we model it in the presence of diverse and previously underexplored non-TTM states, such as talking-to-others, self-talking, and background conditions. To support this new formulation, we construct an Online TTM Dataset consisting of 406 egocentric video clips with approximately 900K annotated frames, each labeled with frame-level social interaction categories (e.g., background, TTM, talking-to-others, self-talking), by extending the Ego4D social interaction benchmark. In this benchmark, we evaluate five adapted baselines and develop a new model that integrates social cues across modalities. Experimental results show that our multimodal model, which jointly leverages audio, visual, and speech-semantic cues, achieves 75.5% frame-level F1 on TTM, outperforming strong baselines and enabling a systematic analysis of how different speaking states affect TTM recognition.
- Open Access
- 24 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-25
What to Distinguish and How? Opportunities and Challenges of Augmenting Multiple, Cluttered Objects in Complex Scenes for People with Low Vision
- 24 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-25
Open AccessWhat to Distinguish and How? Opportunities and Challenges of Augmenting Multiple, Cluttered Objects in Complex Scenes for People with Low Vision
- Open Access
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-26
Touch with Meaning: A Contextual Analysis of Social Touch
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-26
Open AccessTouch with Meaning: A Contextual Analysis of Social Touch
- Open Access
Vision Token Reduction via Attention-Driven Self-Compression for Efficient Multimodal Large Language Models
Multimodal Large Language Models (MLLMs) incur significant computational cost from processing numerous vision tokens through all LLM layers. Prior pruning methods operate either before the LLM, limiting generality due to diverse encoder-projector designs or within the LLM using heuristics that are incompatible with FlashAttention. We take a different approach: rather than identifying unimportant tokens, we treat the LLM itself as the optimal guide for compression. Observing that deeper layers naturally transmit vision-to-text information, we introduce Attention-Driven Self-Compression (ADSC), a simple, broadly applicable method that progressively reduces vision tokens using only the LLM's attention mechanism. Our method applies uniform token downsampling at selected layers, forming bottlenecks that encourage the model to reorganize and compress information into the remaining tokens. It requires no score computation, auxiliary modules, or attention modification, and remains fully compatible with FlashAttention. Applied to LLaVA-1.5, ADSC reduces FLOPs by 53.7% and peak KV-cache memory by 56.7%, while preserving 98.2% of the original model performance. Across multiple benchmarks, it outperforms prior pruning approaches in both efficiency and accuracy. Crucially, under high compression ratios, our method remains robust while heuristic-based techniques degrade sharply.
Open AccessVision Token Reduction via Attention-Driven Self-Compression for Efficient Multimodal Large Language Models
Multimodal Large Language Models (MLLMs) incur significant computational cost from processing numerous vision tokens through all LLM layers. Prior pruning methods operate either before the LLM, limiting generality due to diverse encoder-projector designs or within the LLM using heuristics that are incompatible with FlashAttention. We take a different approach: rather than identifying unimportant tokens, we treat the LLM itself as the optimal guide for compression. Observing that deeper layers naturally transmit vision-to-text information, we introduce Attention-Driven Self-Compression (ADSC), a simple, broadly applicable method that progressively reduces vision tokens using only the LLM's attention mechanism. Our method applies uniform token downsampling at selected layers, forming bottlenecks that encourage the model to reorganize and compress information into the remaining tokens. It requires no score computation, auxiliary modules, or attention modification, and remains fully compatible with FlashAttention. Applied to LLaVA-1.5, ADSC reduces FLOPs by 53.7% and peak KV-cache memory by 56.7%, while preserving 98.2% of the original model performance. Across multiple benchmarks, it outperforms prior pruning approaches in both efficiency and accuracy. Crucially, under high compression ratios, our method remains robust while heuristic-based techniques degrade sharply.
- 6 February 2026
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 134 (3) , 104
High-Quality Sound Separation Across Diverse Categories via Visually-Guided Generative Modeling
- 6 February 2026
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 134 (3) , 104
High-Quality Sound Separation Across Diverse Categories via Visually-Guided Generative Modeling
- Open Access
ARGaze: Autoregressive Transformers for Online Egocentric Gaze Estimation
Online egocentric gaze estimation predicts where a camera wearer is looking from first-person video using only past and current frames, a task essential for augmented reality and assistive technologies. Unlike third-person gaze estimation, this setting lacks explicit head or eye signals, requiring models to infer current visual attention from sparse, indirect cues such as hand-object interactions and salient scene content. We observe that gaze exhibits strong temporal continuity during goal-directed activities: knowing where a person looked recently provides a powerful prior for predicting where they look next. Inspired by vision-conditioned autoregressive decoding in vision-language models, we propose ARGaze, which reformulates gaze estimation as sequential prediction: at each timestep, a transformer decoder predicts current gaze by conditioning on (i) current visual features and (ii) a fixed-length Gaze Context Window of recent gaze target estimates. This design enforces causality and enables bounded-resource streaming inference. We achieve state-of-the-art performance across multiple egocentric benchmarks under online evaluation, with extensive ablations validating that autoregressive modeling with bounded gaze history is critical for robust prediction. We will release our source code and pre-trained models.
Open AccessARGaze: Autoregressive Transformers for Online Egocentric Gaze Estimation
Online egocentric gaze estimation predicts where a camera wearer is looking from first-person video using only past and current frames, a task essential for augmented reality and assistive technologies. Unlike third-person gaze estimation, this setting lacks explicit head or eye signals, requiring models to infer current visual attention from sparse, indirect cues such as hand-object interactions and salient scene content. We observe that gaze exhibits strong temporal continuity during goal-directed activities: knowing where a person looked recently provides a powerful prior for predicting where they look next. Inspired by vision-conditioned autoregressive decoding in vision-language models, we propose ARGaze, which reformulates gaze estimation as sequential prediction: at each timestep, a transformer decoder predicts current gaze by conditioning on (i) current visual features and (ii) a fixed-length Gaze Context Window of recent gaze target estimates. This design enforces causality and enables bounded-resource streaming inference. We achieve state-of-the-art performance across multiple egocentric benchmarks under online evaluation, with extensive ablations validating that autoregressive modeling with bounded gaze history is critical for robust prediction. We will release our source code and pre-trained models.
- Open Access
Explainable AI-Generated Image Detection RewardBench
Conventional, classification-based AI-generated image detection methods cannot explain why an image is considered real or AI-generated in a way a human expert would, which reduces the trustworthiness and persuasiveness of these detection tools for real-world applications. Leveraging Multimodal Large Language Models (MLLMs) has recently become a trending solution to this issue. Further, to evaluate the quality of generated explanations, a common approach is to adopt an "MLLM as a judge" methodology to evaluate explanations generated by other MLLMs. However, how well those MLLMs perform when judging explanations for AI-generated image detection generated by themselves or other MLLMs has not been well studied. We therefore propose \textbf{XAIGID-RewardBench}, the first benchmark designed to evaluate the ability of current MLLMs to judge the quality of explanations about whether an image is real or AI-generated. The benchmark consists of approximately 3,000 annotated triplets sourced from various image generation models and MLLMs as policy models (detectors) to assess the capabilities of current MLLMs as reward models (judges). Our results show that the current best reward model scored 88.76\% on this benchmark (while human inter-annotator agreement reaches 98.30\%), demonstrating that a visible gap remains between the reasoning abilities of today's MLLMs and human-level performance. In addition, we provide an analysis of common pitfalls that these models frequently encounter. Code and benchmark are available at https://github.com/RewardBench/XAIGID-RewardBench.
Open AccessExplainable AI-Generated Image Detection RewardBench
Conventional, classification-based AI-generated image detection methods cannot explain why an image is considered real or AI-generated in a way a human expert would, which reduces the trustworthiness and persuasiveness of these detection tools for real-world applications. Leveraging Multimodal Large Language Models (MLLMs) has recently become a trending solution to this issue. Further, to evaluate the quality of generated explanations, a common approach is to adopt an "MLLM as a judge" methodology to evaluate explanations generated by other MLLMs. However, how well those MLLMs perform when judging explanations for AI-generated image detection generated by themselves or other MLLMs has not been well studied. We therefore propose \textbf{XAIGID-RewardBench}, the first benchmark designed to evaluate the ability of current MLLMs to judge the quality of explanations about whether an image is real or AI-generated. The benchmark consists of approximately 3,000 annotated triplets sourced from various image generation models and MLLMs as policy models (detectors) to assess the capabilities of current MLLMs as reward models (judges). Our results show that the current best reward model scored 88.76\% on this benchmark (while human inter-annotator agreement reaches 98.30\%), demonstrating that a visible gap remains between the reasoning abilities of today's MLLMs and human-level performance. In addition, we provide an analysis of common pitfalls that these models frequently encounter. Code and benchmark are available at https://github.com/RewardBench/XAIGID-RewardBench.
- Open Access
Toward Gaze Target Detection of Young Autistic Children
The automatic detection of gaze targets in autistic children through artificial intelligence can be impactful, especially for those who lack access to a sufficient number of professionals to improve their quality of life. This paper introduces a new, real-world AI application for gaze target detection in autistic children, which predicts a child's point of gaze from an activity image. This task is foundational for building automated systems that can measure joint attention-a core challenge in Autism Spectrum Disorder (ASD). To facilitate the study of this challenging application, we collected the first-ever Autism Gaze Target (AGT) dataset. We further propose a novel Socially Aware Coarse-to-Fine (SACF) gaze detection framework that explicitly leverages the social context of a scene to overcome the class imbalance common in autism datasets-a consequence of autistic children's tendency to show reduced gaze to faces. It utilizes a two-pathway architecture with expert models specialized in social and non-social gaze, guided by a context-awareness gate module. The results of our comprehensive experiments demonstrate that our framework achieves new state-of-the-art performance for gaze target detection in this population, significantly outperforming existing methods, especially on the critical minority class of face-directed gaze.
Open AccessToward Gaze Target Detection of Young Autistic Children
The automatic detection of gaze targets in autistic children through artificial intelligence can be impactful, especially for those who lack access to a sufficient number of professionals to improve their quality of life. This paper introduces a new, real-world AI application for gaze target detection in autistic children, which predicts a child's point of gaze from an activity image. This task is foundational for building automated systems that can measure joint attention-a core challenge in Autism Spectrum Disorder (ASD). To facilitate the study of this challenging application, we collected the first-ever Autism Gaze Target (AGT) dataset. We further propose a novel Socially Aware Coarse-to-Fine (SACF) gaze detection framework that explicitly leverages the social context of a scene to overcome the class imbalance common in autism datasets-a consequence of autistic children's tendency to show reduced gaze to faces. It utilizes a two-pathway architecture with expert models specialized in social and non-social gaze, guided by a context-awareness gate module. The results of our comprehensive experiments demonstrate that our framework achieves new state-of-the-art performance for gaze target detection in this population, significantly outperforming existing methods, especially on the critical minority class of face-directed gaze.
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 10486-10495
AV-DiT: Taming Image Diffusion Transformers for Efficient Joint Audio and Video Generation
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 10486-10495
Open AccessAV-DiT: Taming Image Diffusion Transformers for Efficient Joint Audio and Video Generation
- 19 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2025, 6753-6762
Introduction to the First Workshop on Vision Foundation Models and Generative AI for Accessibility
Vision foundation models and generative AI have recently made remarkable advances, enabling groundbreaking applications across computer vision, multimodal reasoning, and interactive AI systems. These models hold immense potential to enhance accessibility by powering assistive technologies that improve the daily lives of individuals with disabilities. At the same time, applying foundation models and generative AI in accessibility contexts presents key challenges, including fairness and bias, data scarcity, interpretability, and real-world deployment. This CV4A11y workshop brings together researchers from computer vision, human-computer interaction, and accessibility communities to identify challenges, explore emerging opportunities, and foster collaboration toward building inclusive, efficient, and responsible AI-powered accessibility systems.
- 19 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2025, 6753-6762
Introduction to the First Workshop on Vision Foundation Models and Generative AI for Accessibility
Vision foundation models and generative AI have recently made remarkable advances, enabling groundbreaking applications across computer vision, multimodal reasoning, and interactive AI systems. These models hold immense potential to enhance accessibility by powering assistive technologies that improve the daily lives of individuals with disabilities. At the same time, applying foundation models and generative AI in accessibility contexts presents key challenges, including fairness and bias, data scarcity, interpretability, and real-world deployment. This CV4A11y workshop brings together researchers from computer vision, human-computer interaction, and accessibility communities to identify challenges, explore emerging opportunities, and foster collaboration toward building inclusive, efficient, and responsible AI-powered accessibility systems.