Adversarial and focused training of abnormal videos for weakly-supervised anomaly detection

Ping He, Fan Zhang, Gang Li, Huibin Li

Abstract: Highlights•The present study aims to address a long-standing issue in weakly supervised video anomaly detection (WS-VAD), by identifying and highlighting the inter-video data imbalance problem.•In order to address the issue of inter-video data imbalance in WS-VAD, we present a new solution called Adversarial and Focused Training (AFT).•The effectiveness of the proposed approach is demonstrated on UCF-Crime, ShanghaiTech, XD-Violence, and UCSD Peds datasets in both the inter-video data imbalanced experimental setting and the original experimental setting.