Image Classification on OPTIMAL-31
98.66AccuracyGeoMELT
Evaluation Results
| Method | Links | |
|---|---|---|
| GeoMELTBackbone=BEiT-3, zero-shot=true2025.12 | 98.66 | |
| SOTASetting=Transductive, Transductive Inference (TransCLIP)=false2025.06 | 93 | |
| RemoteCLIPBackbone=ViT-L-14, zero-shot=true2023.06 | 90.05 | |
| RemoteCLIPBackbone=ViT-L-14, zero-shot=true2023.06 | 90.05 | |
| T-RemoteCLIPSetting=Transductive, Transductive Inference (TransCLIP)=true2025.06 | 87.8 | |
| T-GeoRSCLIPSetting=Transductive, Transductive Inference (TransCLIP)=true2025.06 | 87.3 | |
| GeoRSCLIPArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 86.5 | |
| SigLIP2Architecture=ViT-L/16, Setting=Zero-shot, Template=DST2026.06 | 85.3 | |
| T-SkyCLIP50Setting=Transductive, Transductive Inference (TransCLIP)=true2025.06 | 85.2 | |
| GeoRSCLIPARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 84.9 | |
| SkyCLIPARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 84.1 | |
| MetaCLIP-2.5BARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 83.8 | |
| SkyCLIPArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 83.5 | |
| MetaCLIP-2.5BArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 83.2 | |
| SigLIP2ARCH=ViT-L/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 83.1 | |
| SigLIPArchitecture=ViT-L/16, Setting=Zero-shot, Template=DST2026.06 | 83.1 | |
| SigLIP2Architecture=ViT-B/16, Setting=Zero-shot, Template=DST2026.06 | 83.1 | |
| T-CLIPSetting=Transductive, Transductive Inference (TransCLIP)=true2025.06 | 82.9 | |
| RemoteCLIPArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 82.5 | |
| CLIP LAION-RSArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 82.4 | |
| RemoteCLIPARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 81.2 | |
| SigLIPARCH=ViT-L/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 81 | |
| SigLIP2ARCH=ViT-B/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 80.9 | |
| MetaCLIP-400MArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 80.9 | |
| CLIPArchitecture=ViT-L/14, Setting=Zero-shot, Template=DST2026.06 | 80.7 | |
| RemoteCLIPBackbone=ViT-L, zero-shot=true2025.12 | 80.65 | |
| GeoRSCLIPARCH=ViT-B/32, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 80.6 | |
| RemoteCLIPARCH=ViT-B/32, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 79.9 | |
| GeoRSCLIPArchitecture=ViT-B/32, Setting=Zero-shot, Template=DST2026.06 | 79.8 | |
| MetaCLIP-400MARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 79.7 | |
| GeoRSCLIPSetting=Transductive, Transductive Inference (TransCLIP)=false2025.06 | 79.6 | |
| SkyCLIP50Setting=Transductive, Transductive Inference (TransCLIP)=false2025.06 | 79.5 | |
| MetaCLIP-400MARCH=ViT-B/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 79.3 | |
| RemoteCLIPBackbone=ViT-B, zero-shot=true2025.12 | 79.03 | |
| CLIP LAION-RSARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 79 | |
| MetaCLIP-2.5BARCH=ViT-B/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 78.8 | |
| CLIPARCH=ViT-L/14, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 78.5 | |
| RemoteCLIPBackbone=ViT-B-32, zero-shot=true2023.06 | 77.96 | |
| RemoteCLIPBackbone=ViT-B-32, zero-shot=true2023.06 | 77.96 | |
| MetaCLIP-400MArchitecture=ViT-B/16, Setting=Zero-shot, Template=DST2026.06 | 77.7 | |
| RemoteCLIPSetting=Transductive, Transductive Inference (TransCLIP)=false2025.06 | 77.6 | |
| RemoteCLIPArchitecture=ViT-B/32, Setting=Zero-shot, Template=DST2026.06 | 77.5 | |
| CLIPBackbone=ViT-L-14, zero-shot=true2023.06 | 76.83 | |
| CLIPBackbone=ViT-L-14, zero-shot=true2023.06 | 76.83 | |
| SigLIPARCH=ViT-B/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 76.5 | |
| MetaCLIP-2.5BArchitecture=ViT-B/16, Setting=Zero-shot, Template=DST2026.06 | 76 | |
| SigLIPArchitecture=ViT-B/16, Setting=Zero-shot, Template=DST2026.06 | 75.9 | |
| CLIPARCH=ViT-B/16, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 74.5 | |
| CLIPBackbone=ViT-L, zero-shot=true2025.12 | 73.92 | |
| CLIPSetting=Transductive, Transductive Inference (TransCLIP)=false2025.06 | 73 | |
| CLIPArchitecture=ViT-B/16, Setting=Zero-shot, Template=DST2026.06 | 72.2 | |
| RemoteCLIPBackbone=ResNet-50, zero-shot=true2023.06 | 70.16 | |
| RemoteCLIPBackbone=ResNet-50, zero-shot=true2023.06 | 70.16 | |
| RemoteCLIPARCH=RN-50, Zero-shot=true, Approach=Meta-prompting (LLaMA-3.1)2026.06 | 69.5 | |
| CLIPBackbone=ViT-B-32, zero-shot=true2023.06 | 68.62 | |
| CLIPBackbone=ViT-B-32, zero-shot=true2023.06 | 68.62 | |
| CLIPBackbone=ViT-B, zero-shot=true2025.12 | 66.13 | |
| CLIPBackbone=ResNet-50, zero-shot=true2023.06 | 64.52 | |
| CLIPBackbone=ResNet-50, zero-shot=true2023.06 | 64.52 | |
| RemoteCLIPArchitecture=RN-50, Setting=Zero-shot, Template=DST2026.06 | 63.8 |