Table 2 Cross-dataset Evaluation.
From: Localization and detection of deepfake videos based on self-blending method
Method | Training Set | Test Set AUC (%) | |||
|---|---|---|---|---|---|
Real | Fake | CDF | DFDCP | DFDC | |
Xception36 | \(\checkmark\) | \(\checkmark\) | 65.5 | - | 59.4 |
F3Net12 | \(\checkmark\) | \(\checkmark\) | 67.8 | - | 57.9 |
Two-branch19 | \(\checkmark\) | \(\checkmark\) | 73.4 | - | 60.8 |
FTCN22 | \(\checkmark\) | \(\checkmark\) | 86.9 | - | 74 |
LipForensics25 | \(\checkmark\) | \(\checkmark\) | 82.4 | 73.5 | - |
Face X-Ray15 | \(\checkmark\) | 80.58 | 71.15 | - | |
Ours | \(\checkmark\) | 87.52 | 72.25 | 80.34 | |
