Text-to-Satellite Image Retrieval on CVG-Text Brisbane
52.58R@1MAPS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MAPSQuery=Joint query2026.06 | 52.58 | 55.33 | |
| Pairwise alignmentQuery=Joint query2026.06 | 50.33 | 53.08 | |
| PMRLQuery=Joint query2026.06 | 49.08 | 52.06 | |
| CrossText2LocQuery=Text only2026.06 | 48.75 | 51.5 | |
| GRAMQuery=Joint query2026.06 | 48.64 | 51.57 | |
| VASTQuery=Joint query2026.06 | 47.21 | 50.46 | |
| CrossText2Loc# Trainable Params=428M2026.04 | 46.08 | 48.67 | |
| OursInternVL# Trainable Params=9M2026.04 | 45.25 | 49 | |
| CrossText2Loc# Trainable Params=428M2026.04 | 43.58 | 47.42 | |
| BLIP# Trainable Params=470M2026.04 | 43 | 46.33 | |
| BLIPQuery=Text only2026.06 | 43 | 46.33 | |
| OursInternVL# Trainable Params=9M2026.04 | 37.17 | 41.83 | |
| BLIP# Trainable Params=470M2026.04 | 34.5 | 38.17 | |
| CLIP-L/14# Trainable Params=428M2026.04 | 34.08 | 37.25 | |
| CLIP-L/14# Trainable Params=428M2026.04 | 32.5 | 34.67 | |
| CLIP-L/14Query=Text only2026.06 | 32.5 | 34.67 | |
| CLIP-B/16Query=Text only2026.06 | 30.83 | 32.67 | |
| BLIPQuery=Image only2026.06 | 27.92 | 29.83 | |
| SigLIP-B/16# Trainable Params=203M2026.04 | 25.58 | 27.42 | |
| CLIP-L/14Query=Image only2026.06 | 25.42 | 27.92 | |
| CLIP-B/16Query=Image only2026.06 | 22.75 | 25.08 | |
| X-VLM# Trainable Params=216M2026.04 | 20.46 | 21.94 | |
| SigLIP-B/16# Trainable Params=203M2026.04 | 19.58 | 22 | |
| X-VLM# Trainable Params=216M2026.04 | 15.67 | 17.6 | |
| ViLT# Trainable Params=87.4M2026.04 | 11 | 14.5 | |
| ViLT# Trainable Params=87.4M2026.04 | 8.67 | 11.75 |