Text-Dependent Speaker Verification on RSR 2015 (part 1)
11EER (Condition I)i-s-vector (BLSTM)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| i-s-vector (BLSTM)Architecture=bidirectional LSTM (BLSTM), Framework=multi-task learning2025.12 | 11 | 172 | 2 | |
| i-s-vector (LSTM)Architecture=unidirectional LSTM, Framework=multi-task learning2025.12 | 17 | 198 | 3 | |
| i-vector + s-vectorvectors=i-vector + s-vector, Fusion=direct concatenation2025.12 | 28 | 113 | 3 | |
| i-vectorvectors=i-vector2025.12 | 35 | 113 | 6 |