Facial Expression Recognition on Ava256-Scan (test)
45UARSwin-T
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Swin-TTraining Dataset=SynHMC-7K, # Params=28.3 M2026.07 | 45 | 47.9 | |
| R-18Training Dataset=SynHMC-7K, # Params=11.7 M2026.07 | 43.1 | 46 | |
| Swin-TTraining Dataset=AffectNet-280K, # Params=28.3 M2026.07 | 36.7 | 38.9 | |
| Swin-BTraining Dataset=AffectNet-280K, # Params=87.8 M2026.07 | 36.3 | 39.2 | |
| EfficientFaceTraining Dataset=AffectNet-280K, # Params=1.3 M2026.07 | 31 | 33 | |
| R-50Training Dataset=AffectNet-280K, # Params=25.6 M2026.07 | 30.7 | 33.1 | |
| R-18Training Dataset=AffectNet-280K, # Params=11.7 M2026.07 | 30.6 | 33.1 | |
| Swin-TTraining Dataset=AffectNet-7K, # Params=28.3 M2026.07 | 30.4 | 31.2 | |
| EfficientFaceTraining Dataset=RAF-DB, # Params=1.3 M2026.07 | 29.2 | 31.7 | |
| APViTTraining Dataset=RAF-DB, # Params=65.2 M2026.07 | 25.4 | 29.1 | |
| R-18Training Dataset=AffectNet-7K, # Params=11.7 M2026.07 | 21.8 | 19.4 |