多轮物理引导的视觉-语言模型用于物理基础的异常检测
Multi-turn Physics-informed Vision-language Model for Physics-grounded Anomaly Detection
AI总结：
本文提出一种多轮物理引导的视觉-语言模型，通过编码物体属性和动态约束提升物理基础异常检测性能，实验显示其在视频级检测中达到96.7%的AUROC，优于现有最佳方法。
摘要：
视觉-语言模型（VLMs）在通用推理方面表现出色，但在物理基础的异常检测中受限，因为动态因果理解至关重要。现有VLMs主要基于外观相关性训练，无法捕捉运动学约束，导致在不规则旋转或违反机械运动的异常检测中表现不佳。我们引入了一种物理引导的指令微调框架，将物体属性、运动范式和动态约束显式编码到结构化提示中。通过多轮对话传递这些物理先验知识，我们的方法将因果推理分解为逐步过程，使模型能够稳健地表示正常和异常动态。在Phys-AD基准测试中，我们的方法在视频级检测中达到96.7%的AUROC，显著优于现有最佳方法（66.9%），并产生更优的因果解释（0.777 LLM分数）。本工作展示了结构化物理先验如何将VLMs转化为可靠的动态异常检测器。
英文摘要：
Vision-Language Models (VLMs) demonstrate strong general-purpose reasoning but remain limited in physics-grounded anomaly detection, where causal understanding of dynamics is essential. Existing VLMs, trained predominantly on appearance-centric correlations, fail to capture kinematic constraints, leading to poor performance on anomalies such as irregular rotations or violated mechanical motions. We introduce a physics-informed instruction tuning framework that explicitly encodes object properties, motion paradigms, and dynamic constraints into structured prompts. By delivering these physical priors through multi-turn dialogues, our method decomposes causal reasoning into incremental steps, enabling robust internal representations of normal and abnormal dynamics. Evaluated on the Phys-AD benchmark, our approach achieves 96.7% AUROC in video-level detection--substantially outperforming prior SOTA (66.9%)--and yields superior causal explanations (0.777 LLM score). This work highlights how structured physics priors can transform VLMs into reliable detectors of dynamic anomalies.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
