AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

GeoBridge: A Semantic-Anchored Multi-View Foundation Model Bridging Images and Text for Geo-Localization

Zixuan Song, Jing Zhang, Di Wang, Zidie Zhou, Wenbin Liu, Haonan Guo, En Wang, Bo Du

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add text semantic anchor to image-image alignment (ablation)GeoBridge · GeoLoc Panorama to SatelliteR@1: 6.43 -> 38.87Improved
add text semantic anchor to image-image alignment (ablation)GeoBridge · GeoLoc Drone to SatelliteR@1: 38.20 -> 45.06Improved
add image-image alignment to text-only alignment (ablation)GeoBridge · GeoLoc Drone to SatelliteR@1: 42.83 -> 45.06Improved
add image-image alignment to text-only alignment (ablation)GeoBridge · GeoLoc Panorama to SatelliteR@1: 35.40 -> 38.87Improved
use GeoBridge instead of Sample4Geo on GeoLoccross-view retrieval · GeoLoc Drone to SatelliteR@1: 27.27 -> 45.05Improved
use GeoBridge instead of DAC on University-1652drone-satellite retrieval · University-1652 Drone to SatelliteR@1: 94.67 -> 95.82Improved
use GeoBridge instead of CGSI on SUES-200 at 150m drone to satellitedrone-satellite retrieval · SUES-200 Drone to Satellite (150m)R@1: 95.95 -> 95.68Hurt
use GeoBridge instead of Panorama-BEV on VIGOR-Crossstreet-satellite retrieval · VIGOR-CrossR@1: 72.19 -> 73.87Improved
use GeoBridge instead of AuxGeo on VIGOR-Same Hitstreet-satellite retrieval · VIGOR-SameHit: 93.78 -> 92.24Hurt
use GeoBridge instead of CLIP-L/14 for text-to-image retrievalcross-modal retrieval · RSIEvalR@10: 74.00 -> 88.00Improved
use GeoBridge instead of CrossText2Loc on text to drone image retrieval from street descriptioncross-modal retrieval · GeoLoc street description to drone imageL@50: 7.76 -> 7.61 (satellite), 7.68 -> 8.76 (drone)Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy