Table 15 Proposed AA-DCN using RAF dataset.
From: Improved facial emotion recognition model based on a novel deep convolutional structure
Label | Emotion | Proposed AA-DCN FER | ||
|---|---|---|---|---|
Precision | Recall | F1-score | ||
0 | Angry | 0.77 | 0.87 | 0.82 |
1 | Disgust | 0.85 | 0.79 | 0.82 |
2 | Fear | 0.81 | 0.74 | 0.77 |
3 | Happy | 0.94 | 0.86 | 0.90 |
4 | Neutral | 0.71 | 0.83 | 0.76 |
5 | Sadness | 0.75 | 0.73 | 0.74 |
6 | Surprise | 0.89 | 0.90 | 0.89 |
Over all accuracy | 0.82 | |||
