Table 5.
Segmentation of the speech-Sakar dataset based on the different features of [53].
| Entry | Feature | Number |
|---|---|---|
| 1 | Baseline | 26 |
| 2 | Bandwidth + formant | 8 |
| 3 | Mel-frequency cepstral coefficients (MFCC) | 84 |
| 4 | Wavelet transform applied to F0 | 182 |
| 5 | Vocal fold | 22 |
| 6 | Tunable Q-factor wavelet transform (TQWT) | 432 |