提出一种严格因果的流式视频异常检测方法，实时性更强且理论可解释。
Strictly Causal Streaming Video Anomaly Detection with a Theoretically-Grounded State-Space Core
- 采用可闭环更新的状态空间模型，每帧仅需常数时间与内存。
- 实测在苹果M3 Pro上每帧延迟仅0.74毫秒，远超1300帧/秒。
- 首次建立衰减谱与检测延迟的理论关系，适合边缘部署场景。
现有基于Mamba型状态空间模型（SSMs）的视频异常检测方法仍依赖帧缓冲或滑动窗口，缺乏对时序记忆与检测延迟之间关系的理论分析，且效率评估仅基于GPU吞吐量。本文提出一种严格因果的流式异常检测器，其固定大小状态在每帧更新中仅需O(1)时间和内存，无前瞻机制亦无片段缓存。核心为带有输入和状态依赖衰减门的对角线线性状态空间递归，通过在冻结视觉主干上进行因果下一个嵌入预测实现自监督训练。我们推导出递归衰减谱与检测延迟及最短可捕获异常长度之间的闭式关系，并在UCSD Ped2和CUHK Avenue数据集上实证验证。基于学习到的基底衰减（57–59帧）预测的稳定延迟上限远高于实测值（1.6和18.4帧），表明事件边界门而非基底衰减主导响应速度。进一步在Apple M3 Pro硬件上直接测量端到端延迟与吞吐量，分别为0.74毫秒和0.77毫秒/帧（>1300 FPS）。未经调优配置下，该方法在Ped2和Avenue上分别取得67.9%和70.2%的帧级AUC，略低于非因果的先前基准。消融实验显示，衰减率、状态大小和门控机制的影响具有数据集规模依赖性：在较小的Ped2上降低准确率，而在较大的Avenue上提升性能。缩小准确率差距并扩展至更大基准是下一步工作。
原文摘要 · Abstract (English)
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 recurrence decay spectrum and both detection delay and the shortest anomaly it can reliably capture, then validate empirically on UCSD Ped2 and CUHK Avenue. The settling delay bound predicted from the learned base decay (57 to 59 frames) sits far above the measured detection delay (1.6 and 18.4 frames), showing that the event boundary gate, not the base decay, governs responsiveness. We further report end to end latency and throughput measured directly on Apple M3 Pro hardware, 0.74 ms and 0.77 ms per frame (over 1300 FPS), rather than simulated GPU numbers. With an untuned initial configuration the method reaches 67.9 percent and 70.2 percent frame level AUC on Ped2 and Avenue, trailing prior non causal SSM baselines in accuracy. Ablations over decay rate, state size, and gating reveal that the gate contribution is dataset size dependent, hurting accuracy on the smaller Ped2 training set but helping on the larger Avenue one. Closing this accuracy gap and extending evaluation to a third, larger benchmark are immediate next steps.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。