Table 5 Changes of speech quality before and after the change
From: Black-box adversarial attacks through speech distortion for speech emotion recognition
Method | α∗ | WER (%) |
|---|---|---|
Original | - | 11.23 |
VTLN | 0.15 | 21.40 |
− 0.15 | 23.84 | |
McAdams | 1.20 | 18.69 |
0.80 | 17.47 | |
MSS | 0.25 | 20.06 |