Spatio-temporal graph-based self-labeling for video anomaly detection

Meng Xing, Zhiyong Feng, Yong Su, Yiming Zhang, Changjae Oh, Valeriya V. Gribova, Vladimir Fedorovich Filaretoy, Deshuang Huang

Abstract: Highlights•A self-labeling mechanism to narrow the gap between the predefined abnormal space and the real one.•Spatio-temporal graph modeling to capture object interactions in VAD.•A novel object-level criterion to evaluate the effectiveness of object interaction modeling.