Video-based Facial Expression Recognition on AFEW
59.16AccuracyVGG13+VGG16+ResNet
Evaluation Results
| Method | Links | |
|---|---|---|
| VGG13+VGG16+ResNetFusion strategy=Multiple models, Fusion level=Score level or Feature level2018.05 | 59.16 | |
| VGG13Fusion strategy=Single model2018.05 | 57.07 | |
| Multiple CNN-RNN and C3DFusion strategy=Multiple models, Fusion level=Score level or Feature level2018.05 | 51.8 | |
| SPDNetBire-layer configuration=4-Bire, Fusion strategy=Single model2018.05 | 46.71 | |
| Baseline (RBF Kernel)Kernel type=RBF, Fusion strategy=Single model2018.05 | 45.95 | |
| Single Best CNN-RNNFusion strategy=Single model2018.05 | 45.43 | |
| Baseline (Poly Kernel)Kernel type=Poly, Fusion strategy=Single model2018.05 | 45.43 | |
| Single Best HoloNetFusion strategy=Single model2018.05 | 44.57 | |
| SPDNetBire-layer configuration=3-Bire, Fusion strategy=Single model2018.05 | 44.09 | |
| SPDNetBire-layer configuration=2-Bire, Fusion strategy=Single model2018.05 | 42.25 | |
| Single Best C3DFusion strategy=Single model2018.05 | 39.69 |