Research Article
[Retracted] Modeling and Simulation of English Speech Optimization Teaching Recognition Based on Intelligent Edge Detection Algorithm
Table 3
COSR results before and after FFPLMS feature set optimization.
| Vocabulary | Characteristic parameters | 0 dB | 5 dB | 10 dB | 15 dB | 20 dB | Average recognition rate |
| 10 words | FFPSS_D | 82.43 | 87.62 | 90.48 | 91.90 | 90.95 | 88.68 | PCA-features1 | 88.57 | 89.52 | 93.33 | 91.95 | 92.38 | 91.15 | FFPSS_D + TEOCC | 82.86 | 90.95 | 92.38 | 94.29 | 94.29 | 90.95 | PCA-features2 | 90.52 | 90.48 | 93.33 | 94.81 | 94.81 | 92.79 |
| 20 words | FFPSS_D | 74.86 | 80.67 | 84.71 | 86.22 | 87.71 | 82.83 | PCA-features1 | 78.10 | 82.86 | 86.90 | 89.29 | 91.43 | 85.72 | FFPSS_D + TEOCC | 79.70 | 83.83 | 86.80 | 89.76 | 89.94 | 86.01 | PCA-features2 | 83.33 | 86.67 | 88.81 | 90.95 | 92.38 | 88.43 |
|
|