Deepfake Detection on DFD (AUC (%), EER)
86.53AUCGenDF
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GenDFCategory=Fine-tuned, Trainable Parameters=0.28M2025.12 | 86.53 | 0.221 | |
| ViT+LoRACategory=Fine-tuned, Trainable Parameters=0.27M2025.12 | 83.42 | — | |
| UIA-ViT*Category=Custom-made, Trainable Parameters=81.83M2025.12 | 81.96 | — | |
| XceptionCategory=Pre-trained, Trainable Parameters=20.81M2025.12 | 81.63 | — | |
| EfficientNet-B4Category=Pre-trained, Trainable Parameters=19M2025.12 | 81.48 | — | |
| RECCE*Category=Custom-made, Trainable Parameters=24.69M2025.12 | 81.19 | — | |
| UCFCategory=Custom-made, Trainable Parameters=46.84M2025.12 | 80.74 | — | |
| F3-NetCategory=Custom-made, Trainable Parameters=20.75M2025.12 | 79.75 | — | |
| ViT*Category=Pre-trained, Trainable Parameters=1.5K2025.12 | 69.03 | 0.356 |