Research Article
Pattern Recognition Methods and Features Selection for Speech Emotion Recognition System
Table 4
Performance results of classifiers for prosodic features.
| | LSP | ANN | k-NN | GMM |
| | Neutral state | 4470
13.3% | 1699
5.1% | 72.5% | 3953
11.8% | 2503
7.5% | 61.2% | 6483
19.3% | 4774
14.2% | 57.6% |
| | Stress | 4790
14.3% | 22629
67.4% | 82.5% | 5307
15.8% | 21825
65.0% | 80.4% | 2777
8.3% | 19556
58.2% | 87.6% |
| | | 48.3% | 93.0% | 80.7% | 42.7% | 89.7 | 76.7% | 70.0% | 80.4% | 70.1% | | | Neutral state | Stress | | Neutral state | Stress | | Neutral state | Stress | |
|
|
