Emotion Recognition on DFEW (test)
55.61UARViT_LSTM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ViT_LSTMStatus=SCIU2024.06 | 55.61 | 69.2 | |
| ViT_LSTMStatus=Original2024.06 | 54.75 | 67.07 | |
| Res50_LSTMStatus=SCIU2024.06 | 54.58 | 66.55 | |
| Former-DFERStatus=SCIU2024.06 | 53.78 | 67.88 | |
| Former-DFERStatus=Original2024.06 | 53.69 | 65.7 | |
| Res50_LSTMStatus=Original2024.06 | 51.83 | 65.53 | |
| Res18_LSTMStatus=SCIU2024.06 | 51.4 | 64.72 | |
| Res18_LSTMStatus=Original2024.06 | 51.32 | 63.85 | |
| VGG16_LSTMStatus=SCIU2024.06 | 49.26 | 61.51 | |
| VGG13_LSTMStatus=SCIU2024.06 | 48.41 | 63.18 | |
| VGG13_LSTMStatus=Original2024.06 | 48.36 | 60.62 | |
| VGG16_LSTMStatus=Original2024.06 | 48.03 | 60.23 | |
| S4DSSL=true, Modality=V, Number of Parameters (M)=101, Pre-training Strategy=A: external pre-trained weights or additional labels2025.05 | 0.668 | 0.7668 | |
| FineCLIPERSSL=true, Modality=T+V, Number of Parameters (M)=N/A, Pre-training Strategy=A: external pre-trained weights or additional labels2025.05 | 0.6598 | 0.7621 | |
| S2DSSL=true, Modality=V, Number of Parameters (M)=N/A, Pre-training Strategy=A: external pre-trained weights or additional labels2025.05 | 0.6545 | 0.7481 | |
| VAEmoSSL=true, Modality=A+V, Number of Parameters (M)=39, Pre-training Strategy=B: trained from scratch on VoxCeleb2 only2025.05 | 0.6402 | 0.7578 | |
| HiCMAE-BSSL=true, Modality=A+V, Number of Parameters (M)=81, Pre-training Strategy=B: trained from scratch on VoxCeleb2 only2025.05 | 0.6376 | 0.7501 | |
| MAE-DFERSSL=true, Modality=V, Number of Parameters (M)=85, Pre-training Strategy=B: trained from scratch on VoxCeleb2 only2025.05 | 0.6341 | 0.7443 | |
| HiCMAE-SSSL=true, Modality=A+V, Number of Parameters (M)=46, Pre-training Strategy=B: trained from scratch on VoxCeleb2 only2025.05 | 0.6305 | 0.7433 | |
| AMHSSL=false, Modality=A+V, Number of Parameters (M)=-, Pre-training Strategy=B: trained from scratch on VoxCeleb2 only2025.05 | 0.5448 | 0.6651 |