Deepfake Attribution on OW-DFA-40 Protocol-3
91.5All ACCCAL (w/o KU)
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| CAL (w/o KU)Knowledge of number of novel forgery types (KU)=unknown2025.12 | 91.5 | 92.1 | 91.7 | 83.2 | 85.4 | 79.6 | 95 | |
| CALKnowledge of number of novel forgery types (KU)=known2025.12 | 91 | 91.8 | 91.6 | 80.3 | 83.9 | 77.3 | 94.8 | |
| CDALKnowledge of number of novel forgery types (KU)=known2025.12 | 87.3 | 88.7 | 83.4 | 74.4 | 78.3 | 68.7 | 91.6 | |
| CPLKnowledge of number of novel forgery types (KU)=known2025.12 | 86.3 | 88.1 | 79.2 | 74.4 | 79.4 | 70.4 | 90.3 | |
| SimGCDKnowledge of number of novel forgery types (KU)=known2025.12 | 85.6 | 86.3 | 82.9 | 78.5 | 79.7 | 72.3 | 90.4 | |
| OwMatchKnowledge of number of novel forgery types (KU)=known2025.12 | 84.3 | 88.4 | 87.9 | 66 | 77.6 | 63.3 | 94.8 | |
| PALGCDKnowledge of number of novel forgery types (KU)=known2025.12 | 83.3 | 82.5 | 79.3 | 63 | 71.4 | 61.7 | 89.5 | |
| LegoGCDKnowledge of number of novel forgery types (KU)=known2025.12 | 82 | 81.9 | 75.3 | 68.6 | 74.8 | 64.8 | 86.8 | |
| LPSKnowledge of number of novel forgery types (KU)=known2025.12 | 80 | 86.9 | 85.9 | 57.8 | 70.5 | 52.5 | 92.5 | |
| ProtoGCDKnowledge of number of novel forgery types (KU)=known2025.12 | 78.2 | 81 | 63.4 | 73.1 | 75.9 | 67.3 | 82.1 |