Research Article
[Retracted] Modeling and Simulation of English Speech Optimization Teaching Recognition Based on Intelligent Edge Detection Algorithm
Table 2
COSR results before and after FFPRLS feature set optimization.
| Vocabulary | Characteristic parameters | 0 dB | 5 dB | 10 dB | 15 dB | 20 dB | Average recognition rate |
| 10 words | FFPSS_D | 79.52 | 86.19 | 88.57 | 89.18 | 90.00 | 86.69 | PCA-features1 | 86.67 | 91.43 | 89.05 | 89.57 | 91.43 | 89.63 | FFPSS_D + TEOCC | 81.43 | 89.05 | 91.90 | 91.43 | 92.38 | 89.24 | PCA-features2 | 87.14 | 90.00 | 91.43 | 91.95 | 92.86 | 90.68 |
| 20 words | FFPSS_D | 73.56 | 80.30 | 83.27 | 86.41 | 87.90 | 82.29 | PCA-features1 | 78.57 | 82.14 | 87.38 | 87.95 | 89.76 | 85.16 | FFPSS_D + TEOCC | 74.67 | 83.27 | 87.55 | 89.39 | 89.94 | 84.96 | PCA-features2 | 80.00 | 85.95 | 89.29 | 89.59 | 91.19 | 87.20 |
|
|