Video Anomaly Detection on HR-ShanghaiTech
87.56Frame-AUCSTG-NF+RPC
Evaluation Results
| Method | Links | |
|---|---|---|
| STG-NF+RPCYear=2026, Type=Ours, score calibration2026.06 | 87.56 | |
| PoseWatchYear=2024, Type=Skeleton transformer2026.06 | 87.23 | |
| DA-Flow+RPCYear=2026, Type=Ours, score calibration2026.06 | 87 | |
| SeeKerYear=2025, Type=Skeleton keypoint density2026.06 | 86.9 | |
| TSGADYear=2024, Type=Skeleton and trajectory2026.06 | 81.77 | |
| TrajRECYear=2024, Type=Skeleton and trajectory2026.06 | 77.9 | |
| Multi-timescale model (Ours)2019.08 | 77.04 | |
| Morais et al.2019.08 | 75.4 | |
| Liu et al.2019.08 | 72.7 | |
| Conv-AE2019.08 | 69.8 |