Image Classification on mini-ImageNet-Red (0% noise)
72.1AccuracyNCR
Evaluation results
| Method | Links | |
|---|---|---|
| NCRBackbone=ResNet-18, Training Protocol=NCR2022.02 | 72.1 | |
| NCR + MixupBackbone=ResNet-18, Training Protocol=NCR, Data Augmentation=Mixup2022.02 | 71.7 | |
| Label smoothingBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 71.2 | |
| BootstrapBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 71.1 | |
| Label smoothing + MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 71.1 | |
| StandardBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 70.9 | |
| StandardBackbone=ResNet-18, Training Protocol=Oracle: Clean Subset2022.02 | 70.9 | |
| MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 70.5 | |
| MixupBackbone=ResNet-18, Training Protocol=Oracle: Clean Subset, Data Augmentation=Mixup2022.02 | 70.5 | |
| Bootstrap + MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 69.9 | |
| MOITBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 64.7 |