FakeAVCeleb
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Deepfake Detection | FakeAVCeleb (test) | Accuracy99.8 | 66 | |
| Audio-visual video forgery detection | FakeAVCeleb | Accuracy99.29 | 41 | |
| Audio-Visual Deepfake Detection | FakeAVCeleb | Accuracy99.91 | 17 | |
| Manipulation detection | FakeAVCeleb (AVG-FV) | AP98.5 | 17 | |
| Manipulation detection | FakeAVCeleb FVFA-WL | AP (%)100 | 17 | |
| Manipulation detection | FakeAVCeleb FVFA-GAN | AP99.9 | 17 | |
| Manipulation detection | FakeAVCeleb (FVFA-FS) | AP100 | 17 | |
| Manipulation detection | FakeAVCeleb FVRA-WL | AP97.8 | 17 | |
| Audio-Visual Deepfake Detection | FakeAVCeleb (test) | ACC (Audio-Visual)97.68 | 16 | |
| Deepfake Detection | FakeAVCeleb Intra-dataset | Accuracy99.2 | 12 | |
| Audio-Visual Deepfake Detection | FakeAVCeleb Original (Trim: X) | AUC100 | 10 | |
| Deepfake Detection | FakeAVCeleb-LS | AUC100 | 9 | |
| Deepfake Detection | FakeAVCeleb | Video-level AUC0.8304 | 9 | |
| Face Forgery Detection | FakeAVCeleb (test) | FS Score98.1 | 9 | |
| Manipulation detection | FakeAVCeleb RVFA | AP93.3 | 8 | |
| Video Forgery Detection | FakeAVCeleb LipGAN-manipulated (test) | AP94.3 | 8 | |
| Audio Generation | FakeAVCeleb | FAD171.52 | 5 | |
| Video Generation | FakeAVCeleb | FID26.88 | 5 | |
| Audio-Visual Deepfake Detection | FakeAVCeleb Trimmed | AUC99.2 | 5 | |
| Face forgery detection | FakeAVCeleb (cross-dataset) | AUC99 | 5 | |
| Multimodal Deepfake Detection | FakeAVCeleb (70%-30% in-domain split) | AUC99.9 | 4 | |
| Multimodal Deepfake Detection | FakeAVCeleb (Leave-one-out) | AUC95.3 | 4 | |
| Audio-Visual Deepfake Detection | FakeAVCeleb zero-shot | EER (zero-shot)5.8 | 4 | |
| Audio-Visual Deepfake Detection | FakeAVCeleb | EER5.8 | 4 |
Showing 24 of 24 rows