Research Article
Developments of Machine Learning Schemes for Dynamic Time-Wrapping-Based Speech Recognition
Table 8
Recognition results of DTW speech recognition with most-matching learning (settings of thresholds:
and
).
| | Index of keywords |
Recognition rates after most-matching learning (unsupervised learning) | | One time | Two times | Five times |
| | Index 1 | 98% | 96% | 98% | | Index 2 | 84% | 88% | 90% | | Index 3 | 72% | 68% | 58% | | Index 4 | 62% | 64% | 54% | | Index 5 | 66% | 66% | 58% |
| | Average | 76.4% | 76.4% | 71.6% |
|
|
