OSMDA: OpenStreetMap-based Domain Adaptation for Remote Sensing VLMs

TL;DR

OSMDA introduces a self-contained domain adaptation method for remote sensing VLMs that leverages OpenStreetMap data and the model's own OCR capabilities, eliminating reliance on costly external annotations or large teacher models.

Contribution

The paper presents OSMDA, a novel framework that enables remote sensing VLMs to adapt to new domains using only the model and publicly available OSM data, without manual labels or external teachers.

Findings

Achieves state-of-the-art results on 10 benchmarks.

Substantially cheaper to train than teacher-dependent methods.

Effective in aligning models with crowd-sourced geographic data.

Abstract

Vision-Language Models (VLMs) adapted to remote sensing rely heavily on domain-specific image-text supervision, yet high-quality annotations for satellite and aerial imagery remain scarce and expensive to produce. Prevailing pseudo-labeling pipelines address this gap by distilling knowledge from large frontier models, but this dependence on large teachers is costly, limits scalability, and caps achievable performance at the ceiling of the teacher. We propose OSMDA: a self-contained domain adaptation framework that eliminates this dependency. Our key insight is that a capable base VLM can serve as its own annotation engine: by pairing aerial images with rendered OpenStreetMap (OSM) tiles, we leverage optical character recognition and chart comprehension capabilities of the model to generate captions enriched by OSM's vast auxiliary metadata. The model is then fine-tuned on the resulting corpus with satellite imagery alone, yielding OSMDA-VLM, a domain-adapted VLM that requires no manual labeling and no stronger external model. We conduct exhaustive evaluations spanning 10 benchmarks across image-text-to-text tasks and comparing against 9 competitive baselines. When equally mixed with real data, our method achieves state-of-the-art results, while being substantially cheaper to train than teacher-dependent alternatives. These results suggest that, given a strong foundation model, alignment with crowd-sourced geographic data is a practical and scalable path towards remote sensing domain adaptation. Dataset and model weights will be made publicly available.