Skip to main content

Advertisement

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