Papers › Distilling Aggregated Knowledge for Weakly-Supervised Video Anomaly Detection

Distilling Aggregated Knowledge for Weakly-Supervised Video Anomaly Detection

5 Jun 2024arXiv:2406.02831archive 2025-07-28

Jash Dalvi, Ali Dabouei, Gunjan Dhanuka, Min Xu

Video anomaly detection aims to develop automated models capable of identifying abnormal events in surveillance videos. The benchmark setup for this task is extremely challenging due to: i) the limited size of the training sets, ii) weak supervision provided in terms of video-level labels, and iii) intrinsic class imbalance induced by the scarcity of abnormal events. In this work, we show that distilling knowledge from aggregated representations of multiple backbones into a single-backbone Student model achieves state-of-the-art performance. In particular, we develop a bi-level distillation approach along with a novel disentangled cross-attention-based feature aggregation network. Our proposed approach, DAKD (Distilling Aggregated Knowledge with Disentangled Attention), demonstrates superior performance compared to existing methods across multiple benchmark datasets. Notably, we achieve significant improvements of 1.36%, 0.78%, and 7.02% on the UCF-Crime, ShanghaiTech, and XD-Violence datasets, respectively.

Paper PDF

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Anomaly Detection Anomaly Detection In Surveillance Videos Video Anomaly Detection Weakly-supervised Video Anomaly Detection

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Anomaly Detection In Surveillance Videos	UCF-Crime	DAKD (Weakly-supervised)	AUC	88.34	#21 of 21	Archive leaderboard	report
Anomaly Detection In Surveillance Videos	XD-Violence	DAKD	AP	85.61	#4 of 17	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections