Research Article
Feature Aggregation with Two-Layer Ensemble Framework for Multilingual Speech Emotion Recognition
Table 11
EV RA results for seven emotions using HDFM as training dataset and Korean KETI dataset as testing dataset.
| | Precision | Recall | F1-score |
| Anger | 0.19 | 0.34 | 0.24 | Disgust | 0.29 | 0.22 | 0.25 | Fear | 0.27 | 0.32 | 0.29 | Happiness | 0.37 | 0.49 | 0.42 | Neutral | 0.21 | 0.05 | 0.07 | Sadness | 0.37 | 0.30 | 0.33 | Surprise | 0.26 | 0.16 | 0.20 |
| Accuracy | | | 0.30 |
| Macro average | 0.28 | 0.27 | 0.26 | Weighted average | 0.30 | 0.30 | 0.28 |
|
|