Image Classification on MedMNIST OrganAMNIST (test)
97.62AccuracyOmniRad small
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| OmniRad smallModel size=small2026.02 | 97.62 | 97.3 | 99.95 | — | |
| DINOv3 baseModel size=base2026.02 | 97.61 | 97.28 | 99.88 | — | |
| Radio DINO baseModel size=base2026.02 | 97.35 | 97.2 | 99.93 | — | |
| OmniRad baseModel size=base2026.02 | 97.14 | 96.98 | 99.89 | — | |
| DINOv3 smallModel size=small2026.02 | 97.06 | 96.83 | 99.87 | — | |
| Radio DINO smallModel size=small2026.02 | 96.83 | 96.47 | 99.92 | — | |
| DINO baseModel size=base2026.02 | 96.79 | 96.39 | 99.87 | — | |
| DINO smallModel size=small2026.02 | 96.45 | 95.93 | 99.88 | — | |
| R-LLM2026.02 | 95.22 | — | 99.78 | — | |
| Resnet18Backbone=Resnet182026.02 | 95.1 | — | 99.8 | — | |
| DINOv2 smallModel size=small2026.02 | 95.07 | 94.55 | 99.83 | — | |
| Resnet50Backbone=Resnet502026.02 | 94.7 | — | 99.8 | — | |
| DINOv2 baseModel size=base2026.02 | 94.56 | 94.19 | 99.76 | — | |
| Med ViT baseModel size=base2026.02 | 94.3 | — | 99.7 | — | |
| DenseNet121Backbone=DenseNet1212026.02 | 93.5 | — | 99.7 | — | |
| Med ViT tinyModel size=tiny2026.02 | 93.1 | — | 99.5 | — | |
| Med ViT smallModel size=small2026.02 | 92.8 | — | 99.6 | — | |
| Wasserstein FIObjective Function=Wasserstein FI-based regularizer, Backbone=ResNet-18, Number of repetitions=10, Input Resolution=28x282025.02 | 91.07 | 7.29 | 98.88 | 0.3996 | |
| ERMObjective Function=Empirical Risk Minimization, Backbone=ResNet-18, Number of repetitions=10, Input Resolution=28x282025.02 | 90.98 | 7.31 | 99.22 | 0.3839 | |
| AutoKeras2026.02 | 90.5 | — | 99.4 | — | |
| Google AutoML Vision2026.02 | 88.6 | — | 99 | — | |
| auto-sklearn2026.02 | 76.2 | — | 96.3 | — |