Table 6 The comparative results proposed methods with other methods on both datasets.
From: Multiscale wavelet attention convolutional network for facial expression recognition
Method | Year | Dataset | Precision | Recall | F1 score | Accuracy |
|---|---|---|---|---|---|---|
DDMAFN | 2024 | FESR | 88.086% | 87.346% | 0.872648 | 87.346% |
MLCL-Net | 2022 | 91.712% | 91.636% | 0.915784 | 91.636% | |
ResEmoteNet | 2024 | 92.768% | 92.520% | 0.924622 | 92.520% | |
MA-Net | 2021 | 95.231% | 95.170% | 0.951586 | 95.170% | |
CDERNet | 2024 | 95.112% | 94.942% | 0.948612 | 94.892% | |
wCA-MCNN (Ours) | - | 88.868% | 87.244% | 0.877576 | 92.438% | |
MsC-wCA-ResNet18 (Ours) | - | 95.974% | 94.702% | 0.952918 | 97.664% | |
DDMAFN | 2024 | KDEF | 87.356% | 90.528% | 0.885360 | 93.258% |
MLCL-Net | 2022 | 82.968% | 82.506% | 0.824844 | 91.914% | |
ResEmoteNet | 2024 | 72.408% | 73.880% | 0.720210 | 83.532% | |
MA-Net | 2021 | 84.708% | 87.476% | 0.856496 | 90.482% | |
CDERNet | 2024 | 85.993% | 86.194% | 0.857028 | 92.059% | |
wCA-MCNN (Ours) | - | 95.074% | 94.968% | 0.949420 | 94.968% | |
MsC-wCA-ResNet18 (Ours) | - | 97.510% | 97.483% | 0.974806 | 97.484% |
