发表机构
School of Computer Science and Technology, University of Chinese Academy of Sciences (UCAS); Institute of Computing Technology (ICT), Chinese Academy of Sciences (CAS); Beijing Academy of Artificial Intelligence (BAAI); School of Artificial Intelligence and Robotics, Hunan University（中国科学院大学计算机科学与技术学院; 中国科学院计算技术研究所; 北京智源人工智能研究院; 湖南大学人工智能与机器人学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
UniTraffic-Agent是第10届AI城市挑战赛第3赛道的MR-CAS解决方案，采用观察-推理-行动-验证工作流，在TAR、FETV、PSI-VQA三项任务中取得相应排名，为交通视频推理提供了新方案。
AI 中文摘要
交通视频理解已成为智能交通领域的重要问题，道路视频为事故、违规行为及车辆与弱势道路使用者间的交互提供直接证据。一个实用系统应能解释交通事件如何发展、为何发生以及相关交互何时发生，但多模态大语言模型（MLLMs）难以实现这一点，因为交通视频包含稀疏事件和多样视角。本文介绍UniTraffic-Agent，它是第10届AI城市挑战赛第3赛道的MR-CAS解决方案，涵盖交通异常推理（TAR）及两项域外评估：针对鱼眼交通事件的FETV和针对行人意图推理的PSI-VQA。UniTraffic-Agent遵循观察-推理-行动-验证工作流，采样带时间戳的视觉证据，在单次请求中推理同一视频片段的所有问题，并通过任务特定的动作适配器转换响应。在官方公开排行榜上，MR-CAS在TAR上排名第16，得分为0.5780；在FETV上排名第2，得分为0.4884；在PSI-VQA上排名第4，得分为64.4161。代码可在该https URL获取。
英文摘要
Traffic video understanding has become an important problem in intelligent transportation, as road videos provide direct evidence for accidents, violations, and interactions between vehicles and vulnerable road users. A useful system should explain how a traffic event develops, why it happens, and when the relevant interaction occurs, yet this remains difficult for multimodal large language models (MLLMs) because traffic videos contain sparse events and varied viewpoints. We introduce UniTraffic-Agent, the MR-CAS solution for Track~3 of the 10th AI City Challenge, which includes Traffic Anomaly Reasoning (TAR) and two out-of-domain evaluations: FETV for fisheye traffic events and PSI-VQA for pedestrian intention reasoning. UniTraffic-Agent follows an observe--reason--act--verify workflow that samples timestamped visual evidence, reasons over all questions from the same clip in one request, and converts responses through task-specific action adapters. On the official Public leaderboards, MR-CAS ranks 16th on TAR with a score of 0.5780, 2nd on FETV with 0.4884, and 4th on PSI-VQA with 64.4161. The code is available at https://github.com/Roclp/UniTraffic-Agent.
CommentsThis paper has been accepted to ECCV 2026 AI City Challenge Workshop
