Mengjingcheng Mo莫梦竟成
Ph.D. Candidate
Chongqing University of Posts and Telecommunications
Research Interests
About
I am a Ph.D. candidate in Computer Science and Technology at Chongqing University of Posts and Telecommunications, advised by Prof. Xinbo Gao (高新波) and Associate Prof. Jiaxu Leng (冷佳旭). My current research focuses on Agentic Anomaly Understanding, with interests in video anomaly understanding, active visual reasoning, and multimodal large language models.
My recent work studies how vision-language agents can actively acquire evidence, reason over anomalous events, and understand challenging real-world scenarios such as aerial anomalies, low-light surveillance, and autonomous-driving corner cases.
Selected Publications
View All →Learning to Watch: Active Video Anomaly Understanding via Interleaved Policy Optimization
Mengjingcheng Mo, Jiaxu Leng, Xinbo Gao*
TL;DR: Learns when and where to watch video evidence for more reliable anomaly reasoning.
Breaking the Continuum: Discrete Distribution Learning for Structural MRI Reconstruction
Tianle Lyu†, Mengjingcheng Mo†, Ting Wen, Zhen Song, Zinan Xiong, Yanjie Zhu*
TL;DR: Models MRI reconstruction as discrete distribution learning to improve structural detail recovery.
Retrieval-Guided Contextual Inference for Training-Free Video Anomaly Detection in Low-Light Scenarios
Mengjingcheng Mo, Jiankang Zheng, Jiaxu Leng, Xinbo Gao*
TL;DR: Uses retrieval-guided context to adapt video anomaly detection under low-light conditions without training.
Linguistic Relative Policy Optimization for Video Anomaly Reasoning
Jiaxu Leng, Jiankang Zheng, Mengjingcheng Mo, Zhanjie Wu, Haosheng Chen, Ji Gan, Xinbo Gao*
TL;DR: Optimizes linguistic reasoning traces for stronger video anomaly understanding.
A2Seek: Towards Reasoning-Centric Benchmark for Aerial Anomaly Understanding
Mengjingcheng Mo, Xinyang Tong, Mingpi Tan, Jiaxu Leng*, Jiankang Zheng, Yiran Liu, Haosheng Chen, Ji Gan, Weisheng Li, Xinbo Gao*
TL;DR: Introduces a benchmark for evaluating reasoning over aerial anomaly evidence.
News
Released A2Seek, an active anomaly-seeking benchmark and project page for aerial anomaly understanding.
Updated the research profile around agentic anomaly understanding, active visual reasoning, and multimodal video analysis.