GeoSearch: Augmenting Worldwide Geolocalization with Web-Scale Reverse Image Search and Image Matching
Tung-Duong Le-Duc, Hoang-Quoc Nguyen-Son, Minh-Son Dao
TLDR
GeoSearch enhances worldwide image geolocalization by integrating web-scale reverse image search and a two-layer filtering mechanism into RAG.
Key contributions
- Proposes GeoSearch, an open-world framework for worldwide image geolocalization.
- Integrates web-scale reverse image search into Retrieval-Augmented Generation (RAG) pipelines.
- Augments LMM prompts with database coordinates and textual evidence from web pages.
- Employs a two-layer filtering mechanism (image matching, confidence gating) to reduce noise.
Why it matters
GeoSearch solves a core limitation in image geolocalization by integrating web-scale reverse image search to handle scenes not in fixed databases. This significantly boosts accuracy, crucial for applications needing precise location from diverse real-world images.
Original Abstract
Worldwide image geolocalization, which aims to predict the GPS coordinates of any image on Earth, remains challenging due to global visual diversity. Recent generative approaches based on Retrieval-Augmented Generation (RAG) and Large Multimodal Models (LMMs) leverage candidates retrieved from fixed databases for reasoning, but often struggle with scenes that are absent from the reference set. In this work, we propose GeoSearch, an open-world geolocation framework that integrates web-scale reverse image search into the RAG pipeline. GeoSearch augments LMM prompts with database-retrieved coordinates and textual evidence extracted from web pages. To mitigate noise from irrelevant content, we introduce a two-layer filtering mechanism consisting of image matching, followed by confidence-based gating. Experiments on standard benchmarks Im2GPS3k and YFCC4k demonstrate the superiority of GeoSearch under leakage-aware evaluation. Our code and data are publicly available to support reproducibility.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.