Deepfake Detection on DeepfakeBench Cross Domain multiple (test)
90.67AUC (CDFv1)SICA
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SICATrain Set=FaceForensics++ (c23), Implementation=This paper2026.02 | 90.67 | 90.68 | 94.36 | 82.41 | 87.63 | 83.02 | 97.4 | 89.45 | |
| EffortTrain Set=FaceForensics++ (c23), Implementation=Re-implementation2026.02 | 88.22 | 87.72 | 94.38 | 81.94 | 87.11 | 79.81 | 97.02 | 88.03 | |
| SPSLTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 81.5 | 76.5 | 81.22 | 70.4 | 74.08 | 64.37 | 94.24 | 77.47 | |
| SRMTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 79.26 | 75.52 | 81.2 | 69.95 | 74.08 | 60.14 | 94.27 | 76.35 | |
| EfficientB4Train Set=FaceForensics++ (c23), Implementation=Original2026.02 | 79.09 | 74.87 | 81.48 | 69.55 | 72.83 | 61.62 | 94.72 | 76.31 | |
| CapsuleTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 79.09 | 74.72 | 68.41 | 64.65 | 65.68 | 64.65 | 90.78 | 72.57 | |
| FWATrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 78.97 | 66.8 | 74.03 | 61.32 | 63.75 | 55.51 | 85.39 | 69.4 | |
| FFDTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 78.4 | 74.35 | 80.24 | 70.29 | 74.26 | 60.56 | 94.5 | 76.09 | |
| CORETrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 77.98 | 74.28 | 80.18 | 70.49 | 73.41 | 60.32 | 94.12 | 75.83 | |
| XceptionTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 77.94 | 73.65 | 81.63 | 70.77 | 73.74 | 62.49 | 93.79 | 76.29 | |
| UCFTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 77.93 | 75.27 | 80.74 | 71.91 | 75.94 | 64.62 | 95.28 | 77.38 | |
| F3NetTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 77.69 | 73.52 | 79.75 | 70.21 | 73.54 | 59.14 | 93.47 | 75.33 | |
| RecceTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 76.77 | 73.19 | 81.19 | 71.33 | 74.19 | 60.95 | 94.46 | 76.01 | |
| CNN-AugTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 74.2 | 70.27 | 64.64 | 63.61 | 61.7 | 59.85 | 87.39 | 68.81 | |
| MesoIncepTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 73.66 | 69.66 | 60.69 | 62.26 | 75.61 | 64.38 | 90.49 | 70.96 | |
| Meso4Train Set=FaceForensics++ (c23), Implementation=Original2026.02 | 73.58 | 60.91 | 54.81 | 55.6 | 59.94 | 56.6 | 71.5 | 61.85 | |
| SiaTrain Set=FaceForensics++ (c23), Implementation=Re-implementation2026.02 | 72.98 | 74.24 | 81.02 | 72.36 | 78.34 | 66.08 | 72.7 | 73.96 | |
| X-rayTrain Set=FaceForensics++ (c23), Implementation=Original2026.02 | 70.93 | 67.86 | 76.55 | 63.26 | 69.42 | 65.53 | 89.89 | 71.92 |