Cross-modal Geo-localization on CVG-Text (New York)
84R@1MAPS
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MAPSQuery=Joint query2026.06 | 84 | 85.42 | |
| Pairwise alignmentQuery=Joint query2026.06 | 79.33 | 80.67 | |
| GRAMQuery=Joint query2026.06 | 78.1 | 79.46 | |
| PMRLQuery=Joint query2026.06 | 77.62 | 79.03 | |
| VASTQuery=Joint query2026.06 | 76.84 | 78.22 | |
| CLIP-L/14Query=Image only2026.06 | 72.75 | 74.75 | |
| BLIPQuery=Image only2026.06 | 71.33 | 73.5 | |
| MAPSQuery=Joint query2026.06 | 67.83 | 70.17 | |
| CLIP-B/16Query=Image only2026.06 | 67.58 | 69.42 | |
| Pairwise alignmentQuery=Joint query2026.06 | 64.42 | 67.25 | |
| GRAMQuery=Joint query2026.06 | 62.81 | 65.73 | |
| PMRLQuery=Joint query2026.06 | 62.46 | 66.14 | |
| CrossText2LocQuery=Text only2026.06 | 62.33 | 65.39 | |
| VASTQuery=Joint query2026.06 | 61.38 | 64.52 | |
| CrossText2Loc# Trainable Params=428M2026.04 | 59.08 | 62 | |
| OursInternVL# Trainable Params=9M2026.04 | 56.92 | 63.33 | |
| BLIP# Trainable Params=470M2026.04 | 52.92 | 55.92 | |
| BLIPQuery=Text only2026.06 | 52.92 | 55.92 | |
| PLANET2026.03 | 52.83 | 55.57 | |
| PLANETQuery=Text only2026.06 | 52.83 | 55.58 | |
| CrossText2Loc2026.03 | 50.33 | 53.07 | |
| CrossText2LocQuery=Text only2026.06 | 50.33 | 53.07 | |
| Long-CLIP2026.03 | 46.42 | 48.25 | |
| CrossText2Loc# Trainable Params=428M2026.04 | 46.25 | 48.75 | |
| OursInternVL# Trainable Params=9M2026.04 | 44.75 | 47.5 | |
| CLIP-L/142026.03 | 35.08 | 37.08 | |
| CLIP-L/14# Trainable Params=428M2026.04 | 35.08 | 37.08 | |
| CLIP-L/14Query=Text only2026.06 | 35.08 | 37.08 | |
| BLIP2026.03 | 34.58 | 37.25 | |
| BLIP# Trainable Params=470M2026.04 | 34.58 | 37.25 | |
| BLIPQuery=Text only2026.06 | 34.58 | 37.25 | |
| EVA2-CLIP-L/142026.03 | 34.08 | 35.67 | |
| RemoteCLIP2026.03 | 33.83 | 36.33 | |
| SigLIP-SO400M2026.03 | 33.5 | 34.83 | |
| CLIP-L/14# Trainable Params=428M2026.04 | 31.5 | 33.58 | |
| CLIP-L/14Query=Text only2026.06 | 31.5 | 33.58 | |
| TAG-CLIP2026.03 | 31.17 | 33.25 | |
| BLIPQuery=Image only2026.06 | 28.67 | 29.75 | |
| CLIP-B/16Query=Text only2026.06 | 27.42 | 29.42 | |
| CLIP-B/162026.03 | 26.67 | 28.17 | |
| CLIP-B/16Query=Text only2026.06 | 26.67 | 28.17 | |
| EVA2-CLIP-B/162026.03 | 25.17 | 26.58 | |
| CLIP-L/14Query=Image only2026.06 | 24.58 | 27 | |
| X2-VLM2026.03 | 23.33 | 25.17 | |
| CLIP-B/16Query=Image only2026.06 | 20.83 | 23.17 | |
| SigLIP-B/16# Trainable Params=203M2026.04 | 20.17 | 21.58 | |
| SigLIP-B/162026.03 | 19.67 | 21.08 | |
| SigLIP-B/16# Trainable Params=203M2026.04 | 19.67 | 21.08 | |
| X-VLM# Trainable Params=216M2026.04 | 16.14 | 17.26 | |
| X-VLM2026.03 | 15.74 | 16.86 | |
| X-VLM# Trainable Params=216M2026.04 | 15.74 | 16.86 | |
| ViLT2026.03 | 11.58 | 15.58 | |
| ViLT# Trainable Params=87.4M2026.04 | 11.58 | 15.58 | |
| ViLT# Trainable Params=87.4M2026.04 | 5.83 | 9.92 |