Listener Deepfake Detection on ListenForge (val)
99.87AUCMANet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MANetModality=Audio-Visual, Retrained=true2026.04 | 99.87 | 98.53 | |
| MANetModality=Audio-Visual, Retraining Status=Proposed method (trained on target dataset)2026.04 | 99.87 | 98.53 | |
| MesoNetModality=Visual, Retrained=true2026.04 | 99.86 | 96.99 | |
| MesoNetModality=Visual, Retraining Status=with retraining2026.04 | 99.86 | 96.99 | |
| AVTFDModality=Audio-Visual, Retrained=true2026.04 | 99.39 | 98.43 | |
| AVTFDModality=Audio-Visual, Retraining Status=with retraining2026.04 | 99.39 | 98.43 | |
| CViTModality=Visual, Retrained=true2026.04 | 97.97 | 97.59 | |
| CViTModality=Visual, Retraining Status=with retraining2026.04 | 97.97 | 97.59 | |
| XceptionModality=Visual, Retrained=true2026.04 | 97.92 | 98.05 | |
| XceptionModality=Visual, Retraining Status=with retraining2026.04 | 97.92 | 98.05 | |
| MRDFModality=Audio-Visual, Retrained=true2026.04 | 91.51 | 88.58 | |
| MRDFModality=Audio-Visual, Retraining Status=with retraining2026.04 | 91.51 | 88.58 | |
| CViTModality=Visual, Retraining Status=without retraining2026.04 | 61.92 | 45.07 | |
| XceptionModality=Visual, Retraining Status=without retraining2026.04 | 60.45 | 58.58 | |
| AVTFDModality=Audio-Visual, Retraining Status=without retraining2026.04 | 58.76 | 55.24 | |
| MRDFModality=Audio-Visual, Retraining Status=without retraining2026.04 | 55.81 | 62.47 | |
| AVADModality=Audio-Visual, Retraining Status=without retraining2026.04 | 47.09 | 37.53 | |
| MesoNetModality=Visual, Retraining Status=without retraining2026.04 | 44.68 | 50.39 |