Deepfake Detection on FaceForensics++ (test)
99.87AUCTALL
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| TALLVenue=ICCV23, Evaluation Setup=intra-dataset2026.01 | 99.87 | 98.65 | — | — | — | — | — | |
| M2F2Venue=CVPR25, Evaluation Setup=intra-dataset2026.01 | 99.34 | 98.79 | — | — | — | — | — | |
| RECCEVenue=CVPR22, Evaluation Setup=intra-dataset2026.01 | 99.32 | 97.06 | — | — | — | — | — | |
| TGNVenue=TIFS24, Evaluation Setup=intra-dataset2026.01 | 99.32 | — | — | — | — | — | — | |
| MultiAttVenue=CVPR21, Evaluation Setup=intra-dataset2026.01 | 99.29 | 97.6 | — | — | — | — | — | |
| DSMTraining Dataset=FF++ [7], Publication Year=20252026.03 | 99 | — | — | — | — | — | — | |
| MAREEvaluation Setup=intra-dataset2026.01 | 98.86 | 96.55 | — | — | — | — | — | |
| HFRTraining Dataset=FF++ [7]2026.03 | 98.42 | — | — | — | — | — | — | |
| SSLTraining Dataset=FF++ [7], Publication Year=20222026.03 | 98.4 | — | — | — | — | — | — | |
| PFGTraining Dataset=FF++ [7], Publication Year=20242026.03 | 98.28 | — | — | — | — | — | — | |
| EFFTTraining Dataset=FF++ [7], Publication Year=20252026.03 | 97.2 | — | — | — | — | — | — | |
| UCFTraining Dataset=FF++ [7], Publication Year=20232026.03 | 97.05 | — | — | — | — | — | — | |
| GM-DFTraining Dataset=FF++ [7], Publication Year=20242026.03 | 96.62 | — | — | — | — | — | — | |
| CORETraining Dataset=FF++ [7], Publication Year=20222026.03 | 96.38 | — | — | — | — | — | — | |
| XceptionTraining Dataset=FF++ [7], Publication Year=20232026.03 | 96.37 | — | — | — | — | — | — | |
| F3NetTraining Dataset=FF++ [7], Publication Year=20232026.03 | 96.35 | — | — | — | — | — | — | |
| D3Training Dataset=FF++ [7], Publication Year=20252026.03 | 96.24 | — | — | — | — | — | — | |
| RecceTraining Dataset=FF++ [7], Publication Year=20222026.03 | 96.21 | — | — | — | — | — | — | |
| SFGDTraining Dataset=FF++ [7], Publication Year=20232026.03 | 95.98 | — | — | — | — | — | — | |
| EfficientB4Training Dataset=FF++ [7], Publication Year=20232026.03 | 95.67 | — | — | — | — | — | — | |
| PBDTraining Dataset=FF++ [7], Publication Year=20212026.03 | 93.1 | — | — | — | — | — | — | |
| SpInShieldPipeline=Black-Box Platform2026.05 | 89.92 | — | — | — | — | — | — | |
| SpInShieldPipeline=Multi-stage Codec2026.05 | 87.53 | — | — | — | — | — | — | |
| SpInShieldPipeline=Cross-Platform2026.05 | 84.44 | — | — | — | — | — | — | |
| SpInShieldPipeline=Heavy Codec2026.05 | 83.28 | — | — | — | — | — | — | |
| SLFPipeline=Black-Box Platform2026.05 | 83.11 | — | — | — | — | — | — | |
| PTFDPipeline=Black-Box Platform2026.05 | 82.97 | — | — | — | — | — | — | |
| FakeSTPipeline=Black-Box Platform2026.05 | 82.38 | — | — | — | — | — | — | |
| NACOPipeline=Black-Box Platform2026.05 | 81.23 | — | — | — | — | — | — | |
| FTCNPipeline=Black-Box Platform2026.05 | 81.02 | — | — | — | — | — | — | |
| AltFrePipeline=Black-Box Platform2026.05 | 80.63 | — | — | — | — | — | — | |
| PTFDPipeline=Multi-stage Codec2026.05 | 78.72 | — | — | — | — | — | — | |
| SLFPipeline=Multi-stage Codec2026.05 | 78.58 | — | — | — | — | — | — | |
| TALLPipeline=Black-Box Platform2026.05 | 78.32 | — | — | — | — | — | — | |
| FakeSTPipeline=Multi-stage Codec2026.05 | 77.53 | — | — | — | — | — | — | |
| NACOPipeline=Multi-stage Codec2026.05 | 77.02 | — | — | — | — | — | — | |
| FTCNPipeline=Multi-stage Codec2026.05 | 76.54 | — | — | — | — | — | — | |
| AltFrePipeline=Multi-stage Codec2026.05 | 76.37 | — | — | — | — | — | — | |
| SLFPipeline=Cross-Platform2026.05 | 76.31 | — | — | — | — | — | — | |
| PTFDPipeline=Cross-Platform2026.05 | 76.27 | — | — | — | — | — | — | |
| DFD-FCGPipeline=Black-Box Platform2026.05 | 76.08 | — | — | — | — | — | — | |
| DFD-FCGPipeline=Cross-Platform2026.05 | 75.61 | — | — | — | — | — | — | |
| FakeSTPipeline=Cross-Platform2026.05 | 75.53 | — | — | — | — | — | — | |
| DFD-FCGPipeline=Heavy Codec2026.05 | 75.38 | — | — | — | — | — | — | |
| DFD-FCGPipeline=Multi-stage Codec2026.05 | 75.27 | — | — | — | — | — | — | |
| NACOPipeline=Cross-Platform2026.05 | 74.43 | — | — | — | — | — | — | |
| FTCNPipeline=Cross-Platform2026.05 | 74.34 | — | — | — | — | — | — | |
| TALLPipeline=Multi-stage Codec2026.05 | 73.99 | — | — | — | — | — | — | |
| AFWPipeline=Black-Box Platform2026.05 | 73.73 | — | — | — | — | — | — | |
| AltFrePipeline=Cross-Platform2026.05 | 73.71 | — | — | — | — | — | — | |
| SLFPipeline=Heavy Codec2026.05 | 72.95 | — | — | — | — | — | — | |
| PTFDPipeline=Heavy Codec2026.05 | 72.81 | — | — | — | — | — | — | |
| FakeSTPipeline=Heavy Codec2026.05 | 71.56 | — | — | — | — | — | — | |
| TALLPipeline=Cross-Platform2026.05 | 71.47 | — | — | — | — | — | — | |
| FTCNPipeline=Heavy Codec2026.05 | 71.17 | — | — | — | — | — | — | |
| VideoMAE(B)Pipeline=Black-Box Platform2026.05 | 70.94 | — | — | — | — | — | — | |
| NACOPipeline=Heavy Codec2026.05 | 70.89 | — | — | — | — | — | — | |
| AltFrePipeline=Heavy Codec2026.05 | 70.66 | — | — | — | — | — | — | |
| AFWPipeline=Multi-stage Codec2026.05 | 69.16 | — | — | — | — | — | — | |
| TALLPipeline=Heavy Codec2026.05 | 67.52 | — | — | — | — | — | — | |
| AFWPipeline=Cross-Platform2026.05 | 66.84 | — | — | — | — | — | — | |
| VideoMAE(B)Pipeline=Multi-stage Codec2026.05 | 65.64 | — | — | — | — | — | — | |
| VideoMAE(B)Pipeline=Cross-Platform2026.05 | 64.24 | — | — | — | — | — | — | |
| AFWPipeline=Heavy Codec2026.05 | 63.71 | — | — | — | — | — | — | |
| VideoMAE(B)Pipeline=Heavy Codec2026.05 | 58.8 | — | — | — | — | — | — | |
| DenseNetAttack Scenario=passcode-controlled clean-label attack, Backbone=DenseNet2025.05 | — | 96.34 | 97.78 | 92.98 | 20 | 853 | 620 | |
| EfficientNetAttack Scenario=passcode-controlled clean-label attack, Backbone=EfficientNet2025.05 | — | 97.32 | 98.89 | 96.46 | 0 | 20 | 20 | |
| F3NetAttack Scenario=passcode-controlled clean-label attack, Backbone=F3Net2025.05 | — | 97.95 | 98.21 | 97.68 | 0 | 5 | 40 | |
| FGAttack Scenario=passcode-controlled clean-label attack, Backbone=FG2025.05 | — | 98.53 | 98.3 | 97.53 | 0 | 101 | 440 | |
| MobilNetAttack Scenario=passcode-controlled clean-label attack, Backbone=MobilNet2025.05 | — | 97.46 | 96.34 | 95.81 | 0 | 156 | 180 | |
| NPRAttack Scenario=passcode-controlled clean-label attack, Backbone=NPR2025.05 | — | 96.33 | 96.47 | 97.37 | 0 | 0 | 0 | |
| ResNetAttack Scenario=passcode-controlled clean-label attack, Backbone=ResNet2025.05 | — | 97.32 | 98.13 | 90.91 | 56 | 106 | 180 | |
| SRMAttack Scenario=passcode-controlled clean-label attack, Backbone=SRM2025.05 | — | 98.56 | 97.93 | 90.1 | 30 | 51 | 640 |