Image Classification on mini-ImageNet-Red (80% noise)
51.2AccuracyNCR
Evaluation Results
| Method | Links | |
|---|---|---|
| NCRBackbone=ResNet-18, Training Protocol=NCR2022.02 | 51.2 | |
| Label smoothingBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 50.2 | |
| StandardBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 49.3 | |
| NCR + MixupBackbone=ResNet-18, Training Protocol=NCR, Data Augmentation=Mixup2022.02 | 48.9 | |
| BootstrapBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 48.8 | |
| MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 48.7 | |
| MixupBackbone=ResNet-18, Training Protocol=Oracle: Clean Subset, Data Augmentation=Mixup2022.02 | 48.7 | |
| StandardBackbone=ResNet-18, Training Protocol=Oracle: Clean Subset2022.02 | 47.5 | |
| Label smoothing + MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 47.3 | |
| MOITBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 45.9 | |
| InstanceGM-SSself-supervised learning=true2022.09 | 44.03 | |
| PropMixself-supervised learning=true2022.09 | 43.42 | |
| Bootstrap + MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 42.2 | |
| ELRBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 41.7 | |
| InstanceGMself-supervised learning=false2022.09 | 39.62 | |
| FaMUSself-supervised learning=false2022.09 | 35.5 | |
| D-MixBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 35.4 | |
| DivideMixself-supervised learning=false2022.09 | 34.5 | |
| MixUpself-supervised learning=false2022.09 | 33.58 | |
| MentorMixself-supervised learning=false2022.09 | 33.46 | |
| CEself-supervised learning=false2022.09 | 29.76 |