Facial Expression Recognition on MAFW (5-fold cross-validation)
48.22UARViT-H FlashLite
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ViT-H FlashLitePretrain=U, Modality=V, Face Align.=×, Params (M)=5212026.06 | 48.22 | 61.45 | |
| ViT-L FlashLitePretrain=U, Modality=V, Face Align.=×, Params (M)=3052026.06 | 46.8 | 60.04 | |
| AVF-MAE++-HPretrain=U+S, Modality=V+A, Face Align.=✓, Params (M)=5212026.06 | 46.05 | 60.24 | |
| ViT-B FlashLitePretrain=U, Modality=V, Face Align.=×, Params (M)=862026.06 | 45.48 | 58.38 | |
| AVF-MAE++-LPretrain=U+S, Modality=V+A, Face Align.=✓, Params (M)=3032026.06 | 45.36 | 59.13 | |
| ViT-B ExactPretrain=U, Modality=V, Face Align.=×, Params (M)=862026.06 | 45.24 | 58.69 | |
| MMA-DFERPretrain=U, Modality=V+A, Face Align.=✓2026.06 | 44.25 | 58.45 | |
| AVF-MAE++-BPretrain=U+S, Modality=V+A, Face Align.=✓, Params (M)=1692026.06 | 43.1 | 57.5 | |
| HiCMAEPretrain=U, Modality=V+A, Face Align.=✓, Params (M)=812026.06 | 42.65 | 56.17 | |
| MAE-DFERPretrain=U, Modality=V, Face Align.=✓, Params (M)=852026.06 | 41.62 | 54.31 | |
| SVFAPPretrain=U, Modality=V, Face Align.=✓, Params (M)=782026.06 | 41.19 | 54.28 | |
| VideoMAE*Pretrain=U, Modality=V, Face Align.=✓, Params (M)=862026.06 | 40.87 | 53.51 | |
| DFER-CLIPPretrain=U, Modality=V+L, Face Align.=✓, Params (M)=1532026.06 | 39.89 | 52.55 | |
| GPT-4oModality=V+A+L2026.06 | 38.29 | 48.82 | |
| EmoCLIPPretrain=U, Modality=V+L2026.06 | 31.41 | 36.18 | |
| C3DModality=V, Params (M)=792026.06 | 31.17 | 42.25 | |
| Former-DFERPretrain=S, Modality=V, Params (M)=182026.06 | 31.16 | 43.27 |