Face Forgery Detection on FaceForensics++ (FF++) c23 light compression
96.69AccuracyFDFL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FDFLdetection_level=image-based2021.03 | 96.69 | 0.993 | 0.985 | |
| Two-branchdetection_level=video-based2021.03 | 96.43 | 0.991 | 0.984 | |
| Xceptiondetection_level=image-based, source=original paper2021.03 | 95.73 | — | — | |
| Xceptiondetection_level=image-based, is_baseline=true2021.03 | 94.24 | 0.972 | 0.903 | |
| MesoNetdetection_level=image-based2021.03 | 83.1 | — | — | |
| Bayar and Stammdetection_level=image-based2021.03 | 82.97 | — | — | |
| Rahmouni et al.detection_level=image-based2021.03 | 79.08 | — | — | |
| Cozzolino et al.detection_level=image-based2021.03 | 78.45 | — | — | |
| Steg. Features + SVMdetection_level=image-based2021.03 | 70.97 | — | — | |
| DSP-FWAdetection_level=image-based2021.03 | — | 0.575 | 0.516 | |
| Face X-raydetection_level=image-based2021.03 | — | 0.874 | — |