Image Classification on ImageNet LT (Robustness and Corruption Evaluation)
42.85Clean AccuracyLA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LABackbone=ResNet182025.01 | 42.85 | 15.87 | 66 | 20.36 | |
| CEBackbone=ResNet182025.01 | 38.31 | 14.8 | 64.67 | 20.21 | |
| HEMBackbone=ResNet182025.01 | 38.25 | 14.4 | 75.4 | 26.33 | |
| HEM-Backbone=ResNet182025.01 | 36.69 | 14.13 | 72.05 | 25.56 | |
| LNBackbone=ResNet182025.01 | 36.31 | 14.99 | 75.73 | 23.38 | |
| LNBackbone=SwinT2025.01 | 33.95 | 9.98 | 63.5 | 18.56 | |
| LNBackbone=ConvNeXt-tiny2025.01 | 33.93 | 13.58 | 54.52 | 15.31 | |
| LABackbone=ConvNeXt-tiny2025.01 | 32.78 | 12.23 | 64.38 | 19.38 | |
| LABackbone=SwinT2025.01 | 32.11 | 8.73 | 62.67 | 19.7 | |
| HEMBackbone=SwinT2025.01 | 30.85 | 8.89 | 81.18 | 23.44 | |
| HEM-Backbone=SwinT2025.01 | 30.84 | 8.5 | 77.08 | 25.44 | |
| CEBackbone=ConvNeXt-tiny2025.01 | 30.46 | 11.4 | 63.51 | 20.72 | |
| CEBackbone=SwinT2025.01 | 29.95 | 8.08 | 55.55 | 19.82 | |
| HEM-Backbone=ConvNeXt-tiny2025.01 | 29.64 | 11.31 | 77.61 | 26.23 | |
| HEMBackbone=ConvNeXt-tiny2025.01 | 29.15 | 10.8 | 79.75 | 23.07 | |
| MMBackbone=ConvNeXt-tiny2025.01 | 22.32 | 8.3 | 65.91 | 20.37 | |
| MMBackbone=ResNet182025.01 | 17.21 | 6.21 | 55.81 | 17.88 | |
| DICEBackbone=ConvNeXt-tiny2025.01 | 4.5 | 2.72 | 44.69 | 14.53 | |
| MMBackbone=SwinT2025.01 | 3.74 | 1.22 | 51.82 | 4.24 | |
| DICEBackbone=ResNet182025.01 | 0.41 | 0.35 | 38.42 | 60.5 | |
| DICEBackbone=SwinT2025.01 | 0.09 | 0.1 | 49.73 | 3.52 |