Geospatial Chain of Thought Reasoning for Enhanced Visual Question Answering on Satellite Imagery
This addresses the need for reliable decision support in high-stakes climate applications like disaster monitoring and urban planning, though it appears incremental as it builds on existing VQA and CoT methods.
The paper tackles the problem of complex geospatial queries in Visual Question Answering (VQA) on satellite imagery by integrating chain of thought (CoT) reasoning with Direct Preference Optimization (DPO), resulting in a 34.9% accuracy improvement over baselines.
Geospatial chain of thought (CoT) reasoning is essential for advancing Visual Question Answering (VQA) on satellite imagery, particularly in climate related applications such as disaster monitoring, infrastructure risk assessment, urban resilience planning, and policy support. Existing VQA models enable scalable interpretation of remote sensing data but often lack the structured reasoning required for complex geospatial queries. We propose a VQA framework that integrates CoT reasoning with Direct Preference Optimization (DPO) to improve interpretability, robustness, and accuracy. By generating intermediate rationales, the model better handles tasks involving detection, classification, spatial relations, and comparative analysis, which are critical for reliable decision support in high stakes climate domains. Experiments show that CoT supervision improves accuracy by 34.9\% over direct baselines, while DPO yields additional gains in accuracy and reasoning quality. The resulting system advances VQA for multispectral Earth observation by enabling richer geospatial reasoning and more effective climate use cases.