SATGround: A Spatially-Aware Approach for Visual Grounding in Remote Sensing

TL;DR

SATGround introduces a novel spatially-aware method for visual grounding in satellite imagery, significantly improving localization accuracy by integrating structured spatial reasoning into vision-language models.

Contribution

It proposes a new structured localization mechanism with a dedicated grounding module, enhancing VLMs' ability to precisely localize objects in satellite scenes.

Findings

33.2% relative improvement over previous methods

Enhanced joint reasoning over language and spatial info

Consistent state-of-the-art performance on benchmarks

Abstract

Vision-language models (VLMs) are emerging as powerful generalist tools for remote sensing, capable of integrating information across diverse tasks and enabling flexible, instruction-based interactions via a chat interface. In this work, we enhance VLM-based visual grounding in satellite imagery by proposing a novel structured localization mechanism. Our approach involves finetuning a pretrained VLM on a diverse set of instruction-following tasks, while interfacing a dedicated grounding module through specialized control tokens for localization. This method facilitates joint reasoning over both language and spatial information, significantly enhancing the model's ability to precisely localize objects in complex satellite scenes. We evaluate our framework on several remote sensing benchmarks, consistently improving the state-of-the-art, including a 33.2% relative improvement over previous methods on visual grounding. Our results highlight the benefits of integrating structured spatial reasoning into VLMs, paving the way for more reliable real-world satellite data analysis. Code will be released upon acceptance.