Frame-level Deepfake Detection on Celeb-DF
0.8317AUCDiffusionFake (En-b4)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DiffusionFake (En-b4)Backbone=EfficientNet-B4, Training Dataset=FF++(HQ)2024.10 | 0.8317 | 0.2459 | |
| UIA-ViTBackbone=ViT, Training Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.8241 | — | |
| DCLTraining Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.823 | 0.2653 | |
| PCL+I2GTraining Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.818 | — | |
| SBITraining Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.8076 | 0.2697 | |
| DiffusionFake (ViT-B)Backbone=ViT-B, Training Dataset=FF++(HQ)2024.10 | 0.8046 | 0.2751 | |
| LRLTraining Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.7826 | 0.2967 | |
| CADDMTraining Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.7756 | 0.3063 | |
| LTWTraining Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.7714 | 0.2934 | |
| GFFTraining Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.7531 | 0.3248 | |
| UCFTraining Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.7527 | — | |
| ViT-BBackbone=ViT-B, Training Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.7464 | 0.3307 | |
| Face X-rayTraining Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.742 | — | |
| EfficientNet-B4 (En-b4)Backbone=EfficientNet-B4, Training Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.7351 | 0.3417 | |
| F3-NetTraining Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.7121 | 0.3403 | |
| MATTraining Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.7065 | 0.3583 | |
| RECCETraining Dataset=FF++(HQ), Reproduced Status=true2024.10 | 0.705 | 0.3534 | |
| XceptionBackbone=Xception, Training Dataset=FF++(HQ), Reproduced Status=false2024.10 | 0.6527 | 0.3877 |