Deepfake Detection on AV-Deepfake1M official (test)
0.999AUCAVH-Align/sup
Evaluation Results
| Method | Links | |
|---|---|---|
| AVH-Align/supModality=AV, access_to_leading_silence_bias=true2024.11 | 0.999 | |
| Silence classifierModality=A, access_to_leading_silence_bias=true2024.11 | 0.9844 | |
| AVH-AlignModality=AV2024.11 | 0.8524 | |
| XceptionModality=V, Method Level=Frame-level2024.11 | 0.6868 | |
| EfficientViTModality=V, Method Level=Frame-level2024.11 | 0.6551 | |
| MesoInception4Modality=V, Method Level=Frame-level2024.11 | 0.6404 | |
| Meso4Modality=V, Method Level=Frame-level2024.11 | 0.6305 | |
| MARLINModality=V, Method Level=Segment-level2024.11 | 0.5803 | |
| MesoInception4Modality=V, Method Level=Segment-level2024.11 | 0.5716 | |
| MDSModality=AV, Method Level=Segment-level2024.11 | 0.5657 | |
| Meso4Modality=V, Method Level=Segment-level2024.11 | 0.5453 |