Video Anomaly Detection on XD-Violence (test)
89.96APLAS-VAD
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| LAS-VADSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 89.96 | — | — | — | — | — | — | — | — | — | — | — | |
| LEC-VADSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 88.47 | — | — | — | — | — | — | — | — | — | — | — | |
| PE-MILSupervision=Weakly Supervised, Modality=RGB+Audio, Feature=I3D+VGGish2026.02 | 88.21 | — | — | — | — | — | — | — | — | — | — | — | |
| PE-MILSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 88.05 | — | — | — | — | — | — | — | — | — | — | — | |
| LAS-VADSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 87.92 | — | — | — | — | — | — | — | — | — | — | — | |
| Holmes-VAUBackbone=ViT, VAD Category=Explainable Multi-modal2024.12 | 87.68 | — | — | — | — | — | — | — | — | — | — | — | |
| Holmes-VAUExplainability=Explainable VAD Methods, Category=Fine-tuning2026.05 | 87.68 | — | — | — | — | — | — | — | — | — | — | — | |
| LEC-VADSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 86.56 | — | — | — | — | — | — | — | — | — | — | — | |
| Ex-VADVenue=ICML’25, Supervision=Weak, Explanation=✓, Training Data=100% / 100%2026.07 | 86.52 | — | — | — | — | — | — | — | — | — | — | — | |
| MemoVADBackbone=VideoMAE-S2026.06 | 85.97 | — | — | — | — | — | — | — | — | — | — | — | |
| TG-MVADVenue=This Work, Modality=Text+RGB+Audio+Flow2026.02 | 85.92 | — | — | — | — | — | — | — | — | — | — | — | |
| ReFLIPSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 85.81 | — | — | — | — | — | — | — | — | — | — | — | |
| OursSupervision=Weakly-supervised, Feature=I3D RGB2023.06 | 85.59 | 0.57 | — | — | — | — | — | — | — | — | — | — | |
| PEL4VADSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 85.59 | — | — | — | — | — | — | — | — | — | — | — | |
| PEL4VADVenue=TIP’24, Supervision=Weak, Explanation=X, Training Data=100% / 100%2026.07 | 85.59 | — | — | — | — | — | — | — | — | — | — | — | |
| ITCSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 85.45 | — | — | — | — | — | — | — | — | — | — | — | |
| π-VADSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 85.37 | — | — | — | — | — | — | — | — | — | — | — | |
| BN-WVADFeature=I3D+VGGish2023.11 | 85.26 | — | — | — | — | — | — | — | — | — | — | — | |
| Holmes-VAD+Explainability=Explainable, Instruction Tuning=false2024.12 | 84.96 | — | — | — | — | — | — | — | — | — | — | — | |
| BN-WVADFeature=I3D2023.11 | 84.93 | — | — | — | — | — | — | — | — | — | — | — | |
| TG-MVADVenue=This Work, Modality=RGB+Audio+Flow2026.02 | 84.78 | — | — | — | — | — | — | — | — | — | — | — | |
| Xiao et al.Venue=ICASSP’23, Modality=RGB+Audio+Flow2026.02 | 84.54 | — | — | — | — | — | — | — | — | — | — | — | |
| VadCLIPBackbone=ViT, VAD Category=Non-explainable2024.12 | 84.51 | — | — | — | — | — | — | — | — | — | — | — | |
| VadCLIPSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 84.51 | — | — | — | — | — | — | — | — | — | — | — | |
| VADCLIPExplainability=Non-explainable VAD Methods, Category=Weakly supervised2026.05 | 84.51 | — | — | — | — | — | — | — | — | — | — | — | |
| VadCLIPBackbone=CLIP2026.06 | 84.51 | — | — | — | — | — | — | — | — | — | — | — | |
| VadCLIPVenue=AAAI’24, Supervision=Weak, Explanation=X, Training Data=100% / 100%2026.07 | 84.51 | — | — | — | — | — | — | — | — | — | — | — | |
| TG-MVADVenue=This Work, Modality=RGB+Audio2026.02 | 84.12 | — | — | — | — | — | — | — | — | — | — | — | |
| Na et al.Venue=MLKE’24, Modality=Text+RGB+Audio+Flow2026.02 | 83.9 | — | — | — | — | — | — | — | — | — | — | — | |
| TG-MVADVenue=This Work, Modality=Text+RGB2026.02 | 83.75 | — | — | — | — | — | — | — | — | — | — | — | |
| TPWNGSupervision=Weakly2024.04 | 83.68 | — | — | — | — | — | — | — | — | — | — | — | |
| Yang et al.Backbone=ViT, VAD Category=Non-explainable2024.12 | 83.68 | — | — | — | — | — | — | — | — | — | — | — | |
| TPWNGSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 83.68 | — | — | — | — | — | — | — | — | — | — | — | |
| TPWNGBackbone=CLIP2026.06 | 83.68 | — | — | — | — | — | — | — | — | — | — | — | |
| SASVenue=arXiv'23, Feature=I3D2023.11 | 83.59 | — | — | — | — | — | — | — | — | — | — | — | |
| CMA-LASupervision=Weakly-supervised, Feature=I3D+VGGish2023.06 | 83.54 | — | — | — | — | — | — | — | — | — | — | — | |
| MACIL-SDVenue=MM'22, Feature=I3D+VGGish2023.11 | 83.4 | — | — | — | — | — | — | — | — | — | — | — | |
| MACIL-SDSupervision=Weakly-supervised, Feature=I3D+VGGish2023.06 | 83.4 | — | — | — | — | — | — | — | — | — | — | — | |
| MACIL-SDVenue=MM’22, Modality=RGB+Audio2026.02 | 83.4 | — | — | — | — | — | — | — | — | — | — | — | |
| MACIL-SDSupervision=Weakly Supervised, Modality=RGB+Audio, Feature=I3D+VGGish2026.02 | 83.4 | — | — | — | — | — | — | — | — | — | — | — | |
| IFS-VADSupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 83.14 | — | — | — | — | — | — | — | — | — | — | — | |
| DMUMode=Weakly-supervised, trained_with_novel_samples=true2023.11 | 82.41 | — | — | — | — | — | — | — | — | — | — | — | |
| DMUSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 82.41 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP-TSABackbone=ViT, VAD Category=Non-explainable2024.12 | 82.19 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP-TSAExplainability=Non-Explainable, Trained on entire training frames=true2024.12 | 82.19 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP-TSASupervision=Weakly Supervised, Modality=RGB, Feature=CLIP2026.02 | 82.19 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP TSAMode=Weakly-supervised, trained_with_novel_samples=true2023.11 | 82.17 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP-TSASupervision=Weakly2024.04 | 82.17 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP-TSAExplainability=Non-explainable VAD Methods, Category=Weakly supervised2026.05 | 82.17 | — | — | — | — | — | — | — | — | — | — | — | |
| CLIP-TSABackbone=CLIP2026.06 | 82.17 | — | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUVenue=AAAI'23, Feature=I3D+VGGish2023.11 | 81.77 | — | — | — | — | — | — | — | — | — | — | — | |
| DMUSupervision=Weakly Supervised, Modality=RGB+Audio, Feature=I3D+VGGish2026.02 | 81.77 | — | — | — | — | — | — | — | — | — | — | — | |
| Pang et al.Venue=ICASSP’21, Modality=RGB+Audio2026.02 | 81.69 | — | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUVenue=AAAI'23, Feature=I3D2023.11 | 81.66 | — | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUSupervision=Weakly-supervised, Feature=I3D RGB2023.06 | 81.66 | 0.65 | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUSupervision=Weakly2024.04 | 81.66 | — | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUBackbone=I3D, VAD Category=Non-explainable2024.12 | 81.66 | — | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUVenue=AAAI’23, Modality=RGB2026.02 | 81.66 | — | — | — | — | — | — | — | — | — | — | — | |
| UR-DMUBackbone=I3D2026.06 | 81.66 | — | — | — | — | — | — | — | — | — | — | — | |
| CU-NetVenue=CVPR'23, Feature=I3D+VGGish2023.11 | 81.43 | — | — | — | — | — | — | — | — | — | — | — | |
| CUPLSupervision=Weakly-supervised, Feature=I3D+VGGish2023.06 | 81.43 | — | — | — | — | — | — | — | — | — | — | — | |
| Zhang et al.Venue=CVPR’23, Modality=RGB+Audio2026.02 | 81.43 | — | — | — | — | — | — | — | — | — | — | — | |
| ECUSupervision=Weakly Supervised, Modality=RGB+Audio, Feature=I3D+VGGish2026.02 | 81.43 | — | — | — | — | — | — | — | — | — | — | — | |
| Cho et al.Supervision=Weakly-supervised, Feature=I3D RGB2023.06 | 81.3 | — | — | — | — | — | — | — | — | — | — | — | |
| CoMoSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 81.3 | — | — | — | — | — | — | — | — | — | — | — | |
| TG-MVADVenue=This Work, Modality=RGB+Flow2026.02 | 80.82 | — | — | — | — | — | — | — | — | — | — | — | |
| DDLSupervision=Weakly-supervised, Feature=I3D RGB2023.06 | 80.72 | — | — | — | — | — | — | — | — | — | — | — | |
| LA-NetSupervision=Weakly Supervised, Modality=RGB, Feature=I3D2026.02 | 80.72 | — | — | — | — | — | — | — | — | — | — | — | |
| MSAFSupervision=Weakly-supervised, Feature=I3D+VGGish2023.06 | 80.51 | — | — | — | — | — | — | — | — | — | — | — | |
| MSAFVenue=SPL’22, Modality=RGB+Audio2026.02 | 80.51 | — | — | — | — | — | — | — | — | — | — | — | |
| S3RVenue=ECCV'22, Feature=I3D2023.11 | 80.26 | — | — | — | — | — | — | — | — | — | — | — | |
| S3RSupervision=Weakly-supervised, Feature=I3D RGB2023.06 | 80.26 | — | — | — | — | — | — | — | — | — | — | — | |
| S3RBackbone=I3D, VAD Category=Non-explainable2024.12 | 80.26 | — | — | — | — | — | — | — | — | — | — | — | |
| S3R+Explainability=Non-Explainable2024.12 | 80.26 | — | — | — | — | — | — | — | — | — | — | — | |
| S3RExplainability=Non-explainable VAD Methods, Category=Weakly supervised2026.05 | 80.26 | — | — | — | — | — | — | — | — | — | — | — | |
| ACFSupervision=Weakly-supervised, Feature=I3D+VGGish2023.06 | 80.13 | — | — | — | — | — | — | — | — | — | — | — | |
| ACFVenue=ICASSP’22, Modality=RGB+Audio2026.02 | 80.13 | — | — | — | — | — | — | — | — | — | — | — | |
| MGFNSupervision=Weakly2024.04 | 80.11 | — | — | — | — | — | — | — | — | — | — | — | |
| MGFN+Explainability=Non-Explainable2024.12 | 80.11 | — | — | — | — | — | — | — | — | — | — | — | |
| MGFNExplainability=Non-explainable VAD Methods, Category=Weakly supervised2026.05 | 80.11 | — | — | — | — | — | — | — | — | — | — | — | |
| MGFNBackbone=I3D2026.06 | 80.11 | — | — | — | — | — | — | — | — | — | — | — | |
| Wu et al.Mode=Weakly-supervised, trained_with_novel_samples=true2023.11 | 80 | — | — | — | — | — | — | — | — | — | — | — | |
| TEVADVenue=CVPRW’23, Modality=Text+RGB2026.02 | 79.8 | — | — | — | — | — | — | — | — | — | — | — | |
| Wu et al.Venue=TMM’22, Modality=RGB+Audio+Flow2026.02 | 79.53 | — | — | — | — | — | — | — | — | — | — | — | |
| Pang et al.Venue=TMM’22, Modality=RGB+Audio2026.02 | 79.37 | — | — | — | — | — | — | — | — | — | — | — | |
| DARVenue=TIFS’23, Modality=RGB+Audio2026.02 | 79.32 | — | — | — | — | — | — | — | — | — | — | — | |
| MGFNBackbone=I3D, VAD Category=Non-explainable2024.12 | 79.19 | — | — | — | — | — | — | — | — | — | — | — | |
| DARVenue=TIFS’23, Modality=RGB+Flow2026.02 | 79.15 | — | — | — | — | — | — | — | — | — | — | — | |
| DARVenue=TIFS’23, Modality=RGB2026.02 | 78.94 | — | — | — | — | — | — | — | — | — | — | — | |
| CU-NetVenue=CVPR'23, Feature=I3D2023.11 | 78.74 | — | — | — | — | — | — | — | — | — | — | — | |
| Zhang et al.Supervision=Weakly2024.04 | 78.74 | — | — | — | — | — | — | — | — | — | — | — | |
| HL-NetVenue=ECCV'20, Feature=I3D+VGGish2023.11 | 78.64 | — | — | — | — | — | — | — | — | — | — | — | |
| Wu et al.Backbone=I3D, VAD Category=Non-explainable2024.12 | 78.64 | — | — | — | — | — | — | — | — | — | — | — | |
| Wu et al.Explainability=Non-Explainable, Trained on entire training frames=true2024.12 | 78.64 | — | — | — | — | — | — | — | — | — | — | — | |
| HL-NetVenue=ECCV’20, Modality=RGB+Audio2026.02 | 78.64 | — | — | — | — | — | — | — | — | — | — | — | |
| Wu et al.Venue=TMM’22, Modality=RGB+Audio2026.02 | 78.64 | — | — | — | — | — | — | — | — | — | — | — | |
| WS-AVVDSupervision=Weakly Supervised, Modality=RGB+Audio, Feature=I3D+VGGish2026.02 | 78.64 | — | — | — | — | — | — | — | — | — | — | — | |
| MSLSupervision=Weakly2024.04 | 78.58 | — | — | — | — | — | — | — | — | — | — | — | |
| MSL+Explainability=Non-Explainable2024.12 | 78.58 | — | — | — | — | — | — | — | — | — | — | — | |
| MSLExplainability=Non-explainable VAD Methods, Category=Weakly supervised2026.05 | 78.58 | — | — | — | — | — | — | — | — | — | — | — |