Table 3.
Numbers of feature vectors and corresponding net-speech durations of the questioned-speaker-condition recordings used for the case-specific validations reported in the present paper.
| Number of feature vectors extracted | Net-speech duration (s) |
|---|---|
| 500 | 5 |
| 1,000 | 10 |
| 1,500 | 15 |
| 2,000 | 20 |
| 3,000 | 30 |
| 4,500 | 45 |
| 6,000 | 60 |
| 9,000 | 90 |
| 12,000 | 120 |
| 18,000 | 180 |