Strictly Causal Streaming Video Anomaly Detection with a Theoretically-Grounded State-Space Core

论文概要

研究领域: AI 作者: Yogesh Kumar 发布时间: 2026-08-25 arXiv: 2608.24810

中文摘要

近期工作将Mamba风格状态空间模型（SSM）应用于视频异常检测，但现有方法仍内部依赖缓冲片段或窗口，缺乏关于时间记忆如何与检测延迟相关的理论解释，且仅通过GPU吞吐量而非这些方法所针对的边缘硬件来基准测试效率。我们引入一种严格因果的流式异常检测器，其固定大小状态在每个传入帧上以O(1)时间和内存更新，无前瞻且无片段缓冲。其时序核心是具有输入和状态依赖衰减门的对角线性状态空间递推，通过冻结视觉骨干上的因果下一嵌入预测进行自监督训练。我们推导了递推衰减谱与检测延迟和它能可靠捕获的最短异常之间的闭式关系。

原文摘要

Recent work has applied Mamba style state space models (SSMs) to video anomaly detection, yet existing approaches still rely on buffering clips or windows internally, lack a theoretical account of how temporal memory relates to detection latency, and benchmark efficiency only through GPU throughput rather than the edge hardware these methods are intended to target. We introduce a strictly causal streaming anomaly detector whose fixed size state is updated in O(1) time and memory per incoming frame, with no lookahead and no clip buffering. Its temporal core is a diagonal linear state space recurrence with an input and state dependent decay gate, trained self supervised through causal next embedding prediction on a frozen visual backbone. We derive a closed form relationship between the recu...

--- *自动采集于 2026-08-27*

#论文 #arXiv #AI #小凯