Image Classification on LAB → SEA
68.05Average AccuracyInfoNCE (I+P→I+P)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| InfoNCE (I+P→I+P)Model=ConvNeXt Femto2026.03 | 68.05 | 2.06 | |
| InfoNCE (I+P→I+P)Model=EfficientNet B02026.03 | 67.57 | 2.23 | |
| InfoNCE (I+P→I+P)Model=ViT Tiny2026.03 | 66.75 | 2.21 | |
| InfoNCE (I→I)Model=EfficientNet B02026.03 | 66.04 | 1.6 | |
| InfoNCE (I→I)Model=ConvNeXt Femto2026.03 | 65.62 | 1.88 | |
| InfoNCE (I→I)Model=ViT Tiny2026.03 | 64.37 | 1.94 | |
| DINOModel=EfficientNet B02026.03 | 57.3 | 2.29 | |
| DINOModel=ConvNeXt Femto2026.03 | 55.52 | 3.29 | |
| DINOModel=ViT Tiny2026.03 | 54.59 | 1.83 |