Fine-grained Understanding on CHOICE img
92AccuracyGeoAlignCLIP
Evaluation results
| Method | Links | |
|---|---|---|
| GeoAlignCLIPVersion=ViT-L/142026.03 | 92 | |
| FG-CLIPVersion=ViT-L/14-3362026.03 | 89.5 | |
| CLIPVersion=ViT-L/142026.03 | 88 | |
| LongCLIPVersion=ViT-L/142026.03 | 88 | |
| GeoAlignCLIPVersion=ViT-B/162026.03 | 88 | |
| FG-CLIPVersion=ViT-B/162026.03 | 87.5 | |
| LRSCLIPVersion=ViT-B/162026.03 | 86.5 | |
| SkyCLIPVersion=ViT-L/14-pct302026.03 | 85 | |
| SkyCLIPVersion=ViT-L/14-pct502026.03 | 84.5 | |
| GeoRSCLIPVersion=ViT-L/142026.03 | 83.5 | |
| GeoRSCLIPVersion=ViT-L/14-3362026.03 | 83.5 | |
| GeoRSCLIPVersion=ViT-B/322026.03 | 83 | |
| LongCLIPVersion=ViT-B/162026.03 | 82.5 | |
| SkyCLIPVersion=ViT-B/322026.03 | 78.5 | |
| RemoteCLIPVersion=ViT-L/142026.03 | 77.5 | |
| RemoteCLIPVersion=ViT-B/322026.03 | 76.5 | |
| VHMVersion=Vicuna-7B2026.03 | 74 | |
| CLIPVersion=ViT-B/162026.03 | 72 | |
| LRSCLIPVersion=ViT-L/142026.03 | 20.2 | |
| GeoChatVersion=Vicuna v1.5-7B2026.03 | 16.5 | |
| SkysenseGPTVersion=Vicuna v1.5-7B2026.03 | 16.5 |