Unbiased multiple instance learning for weakly supervised video anomaly detection

• Main Authors: LYU, Hui, YUE, Zhongqi, SUN, Qianru, LUO, Bin, CUI, Zhen, ZHANG, Hanwang
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2023
• Subjects: Databases and Information Systems; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/8101; https://ink.library.smu.edu.sg/context/sis_research/article/9104/viewcontent/Lv_Unbiased_Multiple_Instance_Learning_for_Weakly_Supervised_Video_Anomaly_Detection_CVPR_2023_paper.pdf
• Institution: Singapore Management University

Weakly Supervised Video Anomaly Detection (WSVAD) is challenging because the binary anomaly label is only given on the video level, but the output requires snippet-level predictions. So, Multiple Instance Learning (MIL) is prevailing in WSVAD. However, MIL is notoriously known to suffer from many false alarms because the snippet-level detector is easily biased towards the abnormal snippets with simple context, confused by the normality with the same bias, and missing the anomaly with a different pattern. To this end, we propose a new MIL framework: Unbiased MIL (UMIL), to learn unbiased anomaly features that improve WSVAD. At each MIL training iteration, we use the current detector to divide the samples into two groups with different context biases: the most confident abnormal/normal snippets and the rest ambiguous ones. Then, by seeking the invariant features across the two sample groups, we can remove the variant context biases.