Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.13628 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908885705555968 |
|---|---|
| author | Yu, Bo Yang, Fengze Liu, Yiming Wang, Chao Luo, Xuewen Li, Taozhe Ke, Ruimin Zhou, Xiaofan Liu, Chenxi |
| author_facet | Yu, Bo Yang, Fengze Liu, Yiming Wang, Chao Luo, Xuewen Li, Taozhe Ke, Ruimin Zhou, Xiaofan Liu, Chenxi |
| contents | The emergence of Vision-Language Models (VLMs) has introduced new paradigms for global image geo-localization through retrieval-augmented generation (RAG) and reasoning-driven inference. However, RAG methods are constrained by retrieval database quality, while reasoning-driven approaches fail to internalize image locatability, relying on inefficient, fixed-depth reasoning paths that increase hallucinations and degrade accuracy. To overcome these limitations, we introduce an Optimized Locatability Score that quantifies an image's suitability for deep reasoning in geo-localization. Using this metric, we curate Geo-ADAPT-51K, a locatability-stratified reasoning dataset enriched with augmented reasoning trajectories for complex visual scenes. Building on this foundation, we propose a two-stage Group Relative Policy Optimization (GRPO) curriculum with customized reward functions that regulate adaptive reasoning depth, visual grounding, and hierarchical geographical accuracy. Our framework, Geo-ADAPT, learns an adaptive reasoning policy, achieves state-of-the-art performance across multiple geo-localization benchmarks, and substantially reduces hallucinations by reasoning both adaptively and efficiently. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_13628 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Locatability-Guided Adaptive Reasoning for Image Geo-Localization with Vision-Language Models Yu, Bo Yang, Fengze Liu, Yiming Wang, Chao Luo, Xuewen Li, Taozhe Ke, Ruimin Zhou, Xiaofan Liu, Chenxi Computer Vision and Pattern Recognition Artificial Intelligence The emergence of Vision-Language Models (VLMs) has introduced new paradigms for global image geo-localization through retrieval-augmented generation (RAG) and reasoning-driven inference. However, RAG methods are constrained by retrieval database quality, while reasoning-driven approaches fail to internalize image locatability, relying on inefficient, fixed-depth reasoning paths that increase hallucinations and degrade accuracy. To overcome these limitations, we introduce an Optimized Locatability Score that quantifies an image's suitability for deep reasoning in geo-localization. Using this metric, we curate Geo-ADAPT-51K, a locatability-stratified reasoning dataset enriched with augmented reasoning trajectories for complex visual scenes. Building on this foundation, we propose a two-stage Group Relative Policy Optimization (GRPO) curriculum with customized reward functions that regulate adaptive reasoning depth, visual grounding, and hierarchical geographical accuracy. Our framework, Geo-ADAPT, learns an adaptive reasoning policy, achieves state-of-the-art performance across multiple geo-localization benchmarks, and substantially reduces hallucinations by reasoning both adaptively and efficiently. |
| title | Locatability-Guided Adaptive Reasoning for Image Geo-Localization with Vision-Language Models |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2603.13628 |