Selected work

Representative Papers

FineVAU: A Novel Human-Aligned Benchmark for Fine-Grained Video Anomaly Understanding

Jan 24, 2026

Current evaluation methods for video anomaly understanding (VAU) struggle to accurately assess models’ fine-grained descriptive capabilities regarding anomalous events, often diverging from human perception. This work reframes VAU as a tripartite parsing task—capturing the anomaly’s “What,” the involved entities “Who,” and the spatial context “Where”—and introduces FineW3, a new benchmark dataset, along with FVScore, a human-aligned evaluation metric. FVScore enables the first interpretable, fine-grained assessment of large vision-language models (LVLMs) based on their coverage of critical visual elements. Structured automatic annotations augment manual labeling, and scoring is grounded in key visual components. Human evaluations demonstrate that FVScore significantly outperforms existing metrics. Experiments further reveal that LVLMs underperform on tasks requiring spatiotemporal fine-grained reasoning but excel in scenarios with static or strong visual cues.

0 citationsRead paper

SortWaste: A Densely Annotated Dataset for Object Detection in Industrial Waste Sorting

Jan 05, 2026arXiv.org

This study addresses the limitations of existing automated sorting systems for industrial waste, which struggle with the high variability, clutter, and visual complexity of real-world waste streams, compounded by a lack of high-quality annotated data from authentic scenarios. To bridge this gap, the authors introduce SortWaste, the first densely annotated object detection dataset specifically designed for industrial waste sorting, along with ClutterScore—a novel metric that quantifies scene-level visual complexity using proxy variables such as object count, category and size entropy, and spatial overlap. Systematic evaluation of state-of-the-art detection models across varying complexity levels reveals significant performance degradation in highly cluttered scenes (e.g., mAP of 59.7% for plastics), highlighting current methodological limitations and establishing a standardized benchmark and toolkit for future research in waste detection.

0 citationsRead paper

Chain-of-Anomaly Thoughts with Large Vision-Language Models

Dec 23, 2025

Large vision-language models (VLMs) suffer from inherent normalcy bias in video surveillance, hindering reliable detection of anomalous behaviors such as criminal acts; moreover, their reasoning lacks inductive bias toward anomalies, leading to systematic false negatives. To address this, we propose CoAT—a multi-agent chain-of-thought anomaly reasoning framework—that introduces, for the first time, an explicit inductive crime bias mechanism at the end of the reasoning chain and designs an anomaly-focused classification layer to counteract normalcy bias. CoAT integrates multi-agent collaborative reasoning, chain-of-thought (CoT) expansion, and anomaly-aware vision-language modeling. Experiments demonstrate that CoAT improves anomaly detection F1-score by 11.8 percentage points on low-resolution videos and boosts anomaly classification accuracy by 3.78 percentage points on high-resolution videos. This work is the first to explicitly embed inductive anomaly bias into the VLM reasoning chain, significantly enhancing sensitivity to and discriminative capability for criminal events.

0 citationsRead paper

Investigating the Experience of Autistic Individuals in Software Engineering

Nov 04, 2025

Traditional research on autistic software engineers centers on deficits and accommodations, neglecting their cognitive strengths and work preferences in engineering practice. Method: Employing sociotechnical grounded theory, this study integrates semi-structured interviews and surveys with thematic analysis guided by a neurodiversity-informed cognitive functioning framework. Contribution/Results: Findings reveal distinct cognitive advantages among autistic engineers—including superior logical reasoning, pattern recognition, sustained attention to detail, and effective human–AI collaboration (especially with AI-assisted tools). They consistently prefer remote work and asynchronous, text-based communication, and demonstrate high intrinsic motivation for technical learning. Critically, this is the first study to systematically map their capabilities in core engineering activities such as code review. The results provide empirical foundations and evidence-based design principles for developing strength-based, occupational inclusion pathways in software engineering.

0 citationsRead paper

Parsley's Group Size Study

Oct 24, 2025

Parsley, a group-based distributed hash table (DHT), suffers from limited stability and scalability in dynamic environments due to the absence of theoretical foundations for its group size parameters—specifically, the hard and soft limits. This work presents the first systematic modeling and analysis of how group size impacts overlay network performance. We propose a dual-bound mechanism comprising a soft target interval and a hard constraint, integrated with preemptive node relocation, dynamic data sharding, and formalized topology operation modeling. Extensive large-scale simulations validate the efficacy of our approach. Compared to state-of-the-art methods, our design significantly reduces group split/merge frequency under high churn, while improving load balancing and robustness. The framework provides both a theoretically grounded parameter configuration methodology and empirically validated guidelines—directly applicable to Parsley and other group-based DHTs.

0 citationsRead paper

Recent publications

Latest Papers

FineVAU: A Novel Human-Aligned Benchmark for Fine-Grained Video Anomaly Understanding

Jan 24, 2026

Current evaluation methods for video anomaly understanding (VAU) struggle to accurately assess models’ fine-grained descriptive capabilities regarding anomalous events, often diverging from human perception. This work reframes VAU as a tripartite parsing task—capturing the anomaly’s “What,” the involved entities “Who,” and the spatial context “Where”—and introduces FineW3, a new benchmark dataset, along with FVScore, a human-aligned evaluation metric. FVScore enables the first interpretable, fine-grained assessment of large vision-language models (LVLMs) based on their coverage of critical visual elements. Structured automatic annotations augment manual labeling, and scoring is grounded in key visual components. Human evaluations demonstrate that FVScore significantly outperforms existing metrics. Experiments further reveal that LVLMs underperform on tasks requiring spatiotemporal fine-grained reasoning but excel in scenarios with static or strong visual cues.

0 citationsRead paper

SortWaste: A Densely Annotated Dataset for Object Detection in Industrial Waste Sorting

Jan 05, 2026arXiv.org

This study addresses the limitations of existing automated sorting systems for industrial waste, which struggle with the high variability, clutter, and visual complexity of real-world waste streams, compounded by a lack of high-quality annotated data from authentic scenarios. To bridge this gap, the authors introduce SortWaste, the first densely annotated object detection dataset specifically designed for industrial waste sorting, along with ClutterScore—a novel metric that quantifies scene-level visual complexity using proxy variables such as object count, category and size entropy, and spatial overlap. Systematic evaluation of state-of-the-art detection models across varying complexity levels reveals significant performance degradation in highly cluttered scenes (e.g., mAP of 59.7% for plastics), highlighting current methodological limitations and establishing a standardized benchmark and toolkit for future research in waste detection.

0 citationsRead paper

Chain-of-Anomaly Thoughts with Large Vision-Language Models

Dec 23, 2025

Large vision-language models (VLMs) suffer from inherent normalcy bias in video surveillance, hindering reliable detection of anomalous behaviors such as criminal acts; moreover, their reasoning lacks inductive bias toward anomalies, leading to systematic false negatives. To address this, we propose CoAT—a multi-agent chain-of-thought anomaly reasoning framework—that introduces, for the first time, an explicit inductive crime bias mechanism at the end of the reasoning chain and designs an anomaly-focused classification layer to counteract normalcy bias. CoAT integrates multi-agent collaborative reasoning, chain-of-thought (CoT) expansion, and anomaly-aware vision-language modeling. Experiments demonstrate that CoAT improves anomaly detection F1-score by 11.8 percentage points on low-resolution videos and boosts anomaly classification accuracy by 3.78 percentage points on high-resolution videos. This work is the first to explicitly embed inductive anomaly bias into the VLM reasoning chain, significantly enhancing sensitivity to and discriminative capability for criminal events.

0 citationsRead paper

Investigating the Experience of Autistic Individuals in Software Engineering

Nov 04, 2025

Traditional research on autistic software engineers centers on deficits and accommodations, neglecting their cognitive strengths and work preferences in engineering practice. Method: Employing sociotechnical grounded theory, this study integrates semi-structured interviews and surveys with thematic analysis guided by a neurodiversity-informed cognitive functioning framework. Contribution/Results: Findings reveal distinct cognitive advantages among autistic engineers—including superior logical reasoning, pattern recognition, sustained attention to detail, and effective human–AI collaboration (especially with AI-assisted tools). They consistently prefer remote work and asynchronous, text-based communication, and demonstrate high intrinsic motivation for technical learning. Critically, this is the first study to systematically map their capabilities in core engineering activities such as code review. The results provide empirical foundations and evidence-based design principles for developing strength-based, occupational inclusion pathways in software engineering.

0 citationsRead paper

Parsley's Group Size Study

Oct 24, 2025

Parsley, a group-based distributed hash table (DHT), suffers from limited stability and scalability in dynamic environments due to the absence of theoretical foundations for its group size parameters—specifically, the hard and soft limits. This work presents the first systematic modeling and analysis of how group size impacts overlay network performance. We propose a dual-bound mechanism comprising a soft target interval and a hard constraint, integrated with preemptive node relocation, dynamic data sharding, and formalized topology operation modeling. Extensive large-scale simulations validate the efficacy of our approach. Compared to state-of-the-art methods, our design significantly reduces group split/merge frequency under high churn, while improving load balancing and robustness. The framework provides both a theoretically grounded parameter configuration methodology and empirically validated guidelines—directly applicable to Parsley and other group-based DHTs.

0 citationsRead paper