Table 2 Analysis of single features across the entire temporal direction.
From: iWAX: interpretable Wav2vec-AASIST-XGBoost framework for voice spoofing detection
Model | Feature # | Dev EER (%) | Eval EER (%) |
|---|---|---|---|
|  | f85 | 0.5602±0.0468 | 2.0992±0.0304 |
XGBoost | f1000 | 0.2176±0.0245 | 0.3612±0.0212 |
|  | f1001 | 0.2742±0.0278 | 0.6388±0.0347 |
