Locatability-Guided Adaptive Reasoning for Image Geo-Localization with Vision-Language Models

TL;DR

This paper introduces a new adaptive reasoning framework for image geo-localization using vision-language models, which improves accuracy and reduces hallucinations by dynamically adjusting reasoning depth based on image locatability.

Contribution

It proposes an Optimized Locatability Score, a new reasoning dataset, and a two-stage adaptive policy for improved geo-localization performance.

Findings

Achieves state-of-the-art results on multiple benchmarks.

Reduces hallucinations in reasoning processes.

Demonstrates effective adaptive reasoning depth control.

Abstract

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.