Semi-supervised Image Classification on CIFAR10-LT gamma=200, 1125 labels (10%) (test)
0.714AccuracyDebiasPL (w/ FixMatch)
Evaluation results
| Method | Links | |
|---|---|---|
| DebiasPL (w/ FixMatch)Backbone=WRN-28-2, Imbalance Ratio (gamma)=2002022.01 | 0.714 | |
| CReST+ w/ DABackbone=WRN-28-2, Imbalance Ratio (gamma)=2002022.01 | 0.677 | |
| CReST w/ DABackbone=WRN-28-2, Imbalance Ratio (gamma)=2002022.01 | 0.641 | |
| FixMatch w/ DA w/ LABackbone=WRN-28-2, Imbalance Ratio (gamma)=2002022.01 | 0.624 | |
| FixMatchBackbone=WRN-28-2, Imbalance Ratio (gamma)=2002022.01 | 0.597 | |
| MixMatchBackbone=WRN-28-2, Imbalance Ratio (gamma)=2002022.01 | 0.545 |