Image Classification on Cifar10-LT exponential
83.66AccuracyOurs(Sum)
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Ours(Sum)Backbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 83.66 | 87.3 | 84.36 | 80.4 | 2.82 | |
| SimSiamBackbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 81.4 | — | — | — | — | |
| FASSLBackbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 80.69 | 86.55 | 76.3 | 78.8 | 4.23 | |
| SDCLRBackbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 80.49 | 88.3 | 78.07 | 75.1 | 5.66 | |
| ByolBackbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 79.33 | 82.63 | 80.56 | 75.95 | 2.79 | |
| SimCLRBackbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 76.77 | 82.16 | 76.9 | 71.2 | 4.47 | |
| MoCoV2Backbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 74.76 | 80.7 | 74.36 | 69.8 | 4.46 | |
| VicRegBackbone=Resnet-18, Imbalance Type=exp, Evaluation Protocol=Linear Evaluation2025.09 | 73.32 | 76.26 | 74.46 | 70.27 | 2.51 |