Image Classification on EuroSAT (Top-1 Accuracy)
97.74Top-1 AccuracyAXIS
Evaluation results
| Method | Links | |
|---|---|---|
| AXISArchitecture=ViT-B-322025.08 | 97.74 | |
| aTLASArchitecture=ViT-B-322025.08 | 95.67 | |
| MaPLe + OursTest-time prompt tuning=True2026.04 | 51.69 | |
| OursTest-time prompt tuning=True2026.04 | 50.62 | |
| SLIPPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.03 | 11.5 | |
| SigLIPPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.03 | 10.2 | |
| FLAIRPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.03 | 7.2 | |
| CLIPPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.03 | 5.7 | |
| ITO sub2Pre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.03 | 4.2 | |
| ITOPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.03 | 1.2 |