Fine-grained Understanding on CHOICE ins
63.8Acc@1VHM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| VHMVersion=Vicuna-7B2026.03 | 63.8 | — | |
| GeoAlignCLIPVersion=ViT-B/162026.03 | 62 | 98.6 | |
| GeoAlignCLIPVersion=ViT-L/142026.03 | 62 | 98.8 | |
| RemoteCLIPVersion=ViT-L/142026.03 | 61.4 | 98.8 | |
| RemoteCLIPVersion=ViT-B/322026.03 | 61.2 | 97.6 | |
| CLIPVersion=ViT-B/162026.03 | 50 | 91.6 | |
| SkyCLIPVersion=ViT-L/14-pct302026.03 | 49.8 | 92.4 | |
| SkyCLIPVersion=ViT-L/14-pct502026.03 | 48.2 | 91.4 | |
| GeoChatVersion=Vicuna v1.5-7B2026.03 | 43.2 | — | |
| SkysenseGPTVersion=Vicuna v1.5-7B2026.03 | 41.6 | — | |
| LongCLIPVersion=ViT-L/142026.03 | 35.8 | 91.6 | |
| CLIPVersion=ViT-L/142026.03 | 32.8 | 86.2 | |
| LongCLIPVersion=ViT-B/162026.03 | 27.8 | 90.8 | |
| LRSCLIPVersion=ViT-L/142026.03 | 20.2 | 86 | |
| FG-CLIPVersion=ViT-L/14-3362026.03 | 18.4 | 89 | |
| FG-CLIPVersion=ViT-B/162026.03 | 13.6 | 87.6 | |
| SkyCLIPVersion=ViT-B/322026.03 | 12.4 | 84.6 | |
| GeoRSCLIPVersion=ViT-L/14-3362026.03 | 4.4 | 61.6 | |
| GeoRSCLIPVersion=ViT-B/322026.03 | 4 | 58.6 | |
| LRSCLIPVersion=ViT-B/162026.03 | 2.8 | 59 | |
| GeoRSCLIPVersion=ViT-L/142026.03 | 2.6 | 58.2 |