Skin Lesion Classification on ISIC 3-way 1-shot 2018 (test)
64.15AccuracyHybrid distil
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Hybrid distilBackbone=ResNet-122021.10 | 64.15 | 61.01 | |
| MoCo_v1 (baseline)Backbone=ResNet-122021.10 | 61.9 | 56.3 | |
| MoCo_v2Backbone=ResNet-122021.10 | 59.95 | 55.98 | |
| SRL-distilBackbone=ResNet-122021.10 | 55.43 | 51.18 | |
| Hybrid distilBackbone=4 conv blocks2021.10 | 54.72 | 47.28 | |
| SRL-simpleBackbone=ResNet-122021.10 | 54.45 | 51.02 | |
| MoCo_v1 (baseline)Backbone=4 conv blocks2021.10 | 53.36 | 46.9 | |
| SimCLRBackbone=ResNet-122021.10 | 52.43 | 44.7 | |
| SRL-distilBackbone=4 conv blocks2021.10 | 50.14 | 43.76 | |
| DAMLBackbone=ResNet-122021.10 | 50.05 | 41.65 | |
| MoCo_v2Backbone=4 conv blocks2021.10 | 48.88 | 44.45 | |
| SimCLRBackbone=4 conv blocks2021.10 | 48.07 | 44.91 | |
| MAMLBackbone=ResNet-122021.10 | 47.49 | 42.33 | |
| DAMLBackbone=4 conv blocks2021.10 | 46.7 | 40.96 | |
| MAMLBackbone=4 conv blocks2021.10 | 46.2 | 39.84 | |
| RelationNetBackbone=ResNet-122021.10 | 46.1 | 39.98 | |
| UMTRABackbone=ResNet-122021.10 | 45.88 | 41.44 | |
| SRL-simpleBackbone=4 conv blocks2021.10 | 45.08 | 37.15 | |
| CACTUS-MAMLBackbone=ResNet-122021.10 | 42.98 | 35.38 | |
| CACTUS-ProtoNetsBackbone=ResNet-122021.10 | 42.67 | 39.24 | |
| UMTRABackbone=4 conv blocks2021.10 | 41.45 | 35.88 | |
| RelationNetBackbone=4 conv blocks2021.10 | 41.11 | 36.47 | |
| CACTUS-MAMLBackbone=4 conv blocks2021.10 | 39.45 | 33.09 | |
| Training from scratchBackbone=ResNet-122021.10 | 37.74 | 29.9 | |
| Training from scratchBackbone=4 conv blocks2021.10 | 35.91 | 30.11 | |
| ProtoNetsBackbone=ResNet-122021.10 | 35.18 | 30.81 | |
| ProtoNetsBackbone=4 conv blocks2021.10 | 34.97 | 28.31 | |
| CACTUS-ProtoNetsBackbone=4 conv blocks2021.10 | 33.24 | 29.69 |