Deepfake detection on ForgeryNet (val)
0.9425AUCMINTIME-XC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MINTIME-XCIdentities=22022.11 | 0.9425 | 0.8764 | |
| MINTIME-XCIdentities=32022.11 | 0.941 | 0.8698 | |
| SlowFast R-50Identities=1, Source=Taken from [26]2022.11 | 0.9388 | 0.8878 | |
| X3D-MIdentities=1, Source=Taken from [26]2022.11 | 0.9375 | 0.8793 | |
| MINTIME-XCIdentities=12022.11 | 0.932 | 0.8596 | |
| SlowFast R-50Identities=1, Trained in current setup=true2022.11 | 0.9086 | 0.8259 | |
| MINTIME-EFIdentities=22022.11 | 0.9045 | 0.8228 | |
| MINTIME-EFIdentities=32022.11 | 0.9028 | 0.8205 | |
| MINTIME-EFIdentities=12022.11 | 0.9013 | 0.8192 | |
| RealForensicsTraining dataset=FF++2022.01 | 0.718 | — | |
| LipForensicsTraining dataset=FF++2022.01 | 0.667 | — | |
| FTCNTraining dataset=FF++2022.01 | 0.573 | — |