Table 3 Different division ways of full-band and the corresponding recognition performance
From: A sub-band-based feature reconstruction approach for robust speaker recognition
Channel | Sub-bands | Recognition (%) |
|---|---|---|
number | ||
12 | 1-2, 3-4, 5-6, 7-8, 9-10, 11-12,13-14,15-16, | 74.18 |
17-18, 19-20, 21-22, 23-24 | ||
8 | 1-3, 4-6, 7-9, 10-12, 13-15, 16-18, 19-21, 22-24 | 78.15 |
6 | 1-4, 5-8, 9-12, 13-16, 17-20, 21-24 | 77.83 |
4 | 1-6, 7-12, 13-18, 19-24 | 84.86 |
3 | 1-8, 9-16, 17-24 | 81.78 |
2 | 1-12, 13-24 | 82.96 |