Audio Deepfake Detection on ASV 2019 (test)
98AccuracyRawformer
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Rawformer2026.01 | 98 | 99 | |
| SALMONN MultiEvaluation Prompt=Multi-prompt2026.01 | 98 | 98 | |
| RawGAT-ST2026.01 | 97 | 93 | |
| Rawnet22026.01 | 93 | 92 | |
| RNNInput Features=STFT & LF2026.01 | 92 | 91 | |
| CNNInput Features=STFT & LF2026.01 | 88 | 90 | |
| CRNNInput Features=STFT & LF2026.01 | 88 | 90 | |
| ConvNeXt-TinyInput Features=STFT & LF2026.01 | 88 | 90 | |
| Whisper, CNNInput Features=Raw Audio, STFT & LF2026.01 | 87 | 89 | |
| Seamless+MLPInput Features=Raw Audio2026.01 | 86 | 88 | |
| Whisper, ConvNeXt-TinyInput Features=Raw Audio, STFT & LF2026.01 | 86 | 88 | |
| Whisper+MLPInput Features=Raw Audio2026.01 | 85 | 88 | |
| Swin TInput Features=STFT & LF2026.01 | 84 | 87 | |
| SinC-CNNInput Features=Raw audio2026.01 | 84 | 87 | |
| Speechbrain+MLPInput Features=Raw Audio2026.01 | 77 | 81 | |
| Pyannote+MLPInput Features=Raw Audio2026.01 | 64 | 71 |