Image Classification on ImageNet-1K (Accuracy & Data Reduction)
80AccuracyFull Training
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Full TrainingBackbone=Swin, Architecture Type=Transformer2025.03 | 80 | — | |
| SeTaBackbone=Swin, Architecture Type=Transformer2025.03 | 80 | 46.4 | |
| SeTaBackbone=Swin, Architecture Type=Transformer2025.03 | 79.3 | 55.5 | |
| SeTaBackbone=Swin, Architecture Type=Transformer2025.03 | 79.2 | 63.2 | |
| Full TrainingBackbone=ResNet-50, Architecture Type=CNN2025.03 | 78.6 | — | |
| SeTaBackbone=ResNet-50, Architecture Type=CNN2025.03 | 78.5 | 30 | |
| SeTaBackbone=ResNet-50, Architecture Type=CNN2025.03 | 78.1 | 45.5 | |
| SeTaBackbone=ResNet-50, Architecture Type=CNN2025.03 | 77.6 | 52.7 | |
| SeTaBackbone=EfficientNet, Architecture Type=CNN2025.03 | 76.3 | 30 | |
| Full TrainingBackbone=EfficientNet, Architecture Type=CNN2025.03 | 76.1 | — | |
| Full TrainingBackbone=Vim, Architecture Type=Mamba2025.03 | 75.7 | — | |
| SeTaBackbone=EfficientNet, Architecture Type=CNN2025.03 | 75.7 | 45 | |
| SeTaBackbone=Vim, Architecture Type=Mamba2025.03 | 75.7 | 30 | |
| SeTaBackbone=EfficientNet, Architecture Type=CNN2025.03 | 75.6 | 52.3 | |
| SeTaBackbone=Vim, Architecture Type=Mamba2025.03 | 74.9 | 44.8 | |
| SeTaBackbone=Vim, Architecture Type=Mamba2025.03 | 74.1 | 55.8 | |
| Full TrainingBackbone=ViT, Architecture Type=Transformer2025.03 | 73.3 | — | |
| SeTaBackbone=ViT, Architecture Type=Transformer2025.03 | 73.3 | 24.6 | |
| SeTaBackbone=ViT, Architecture Type=Transformer2025.03 | 72.1 | 38.3 | |
| SeTaBackbone=ViT, Architecture Type=Transformer2025.03 | 72 | 44.8 | |
| SeTaBackbone=ResNet-18, Architecture Type=CNN2025.03 | 69.6 | 36.1 | |
| Full TrainingBackbone=ResNet-18, Architecture Type=CNN2025.03 | 69.5 | — | |
| SeTaBackbone=ResNet-18, Architecture Type=CNN2025.03 | 68.5 | 44.2 | |
| SeTaBackbone=ResNet-18, Architecture Type=CNN2025.03 | 68.2 | 55.3 | |
| Full TrainingBackbone=MobileNetV3, Architecture Type=CNN2025.03 | 52.2 | — | |
| SeTaBackbone=MobileNetV3, Architecture Type=CNN2025.03 | 52.2 | 26.2 | |
| SeTaBackbone=MobileNetV3, Architecture Type=CNN2025.03 | 52 | 42.8 | |
| SeTaBackbone=MobileNetV3, Architecture Type=CNN2025.03 | 51.3 | 50.8 |