Table 8 Comprehensive evaluation of localization performance using AUC and mIoU metrics for all models on the FF++ dataset with two levels of video quality
From: Multi-attention-based approach for deepfake face and expression swap detection and localization
Models | FF++ (C-23) | FF++ (C-40) | ||
|---|---|---|---|---|
mIoU (%) | AUC (%) | mIoU (%) | AUC (%) | |
Model-A | 85.67 | 95.06 | 82.21 | 93.62 |
Model-B | 85.79 | 95.94 | 80.01 | 91.20 |
Model-C | 80.42 | 93.25 | 74.31 | 90.57 |