Image Classification on ImageNet Hard
23.4AccuracyDeiT-L
Evaluation results
| Method | Links | |
|---|---|---|
| DeiT-LForAug=True2025.03 | 23.4 | |
| DeiT-BForAug=False2025.03 | 22.6 | |
| DeiT-BForAug=True2025.03 | 22.6 | |
| ViT-BForAug=True2025.03 | 22 | |
| DeiT-LForAug=False2025.03 | 21.2 | |
| ViT-SForAug=True2025.03 | 21 | |
| ViT-LForAug=True2025.03 | 20.6 | |
| Swin-SForAug=True2025.03 | 20.5 | |
| ResNet101ForAug=True2025.03 | 19.9 | |
| DeiT-SForAug=False2025.03 | 19.5 | |
| DeiT-SForAug=True2025.03 | 18.5 | |
| Swin-TiForAug=True2025.03 | 18.3 | |
| Swin-SForAug=False2025.03 | 18.2 | |
| ResNet101ForAug=False2025.03 | 18.2 | |
| ViT-SForAug=False2025.03 | 18.1 | |
| ResNet50ForAug=True2025.03 | 17.2 | |
| ViT-BForAug=False2025.03 | 17 | |
| Swin-TiForAug=False2025.03 | 16.2 | |
| ResNet50ForAug=False2025.03 | 16.1 | |
| ViT-LForAug=False2025.03 | 15.6 |