让卫星图像问答更懂地理逻辑，提升气候决策可靠性
Geospatial Chain of Thought Reasoning for Enhanced Visual Question Answering on Satellite Imagery
- 引入思维链推理+直接偏好优化，增强模型解释性
- 复杂地理任务准确率比基线提升34.9%
- 适合灾害监测、城市规划等高风险气候应用
地理空间思维链（Geospatial Chain of Thought, CoT）推理对推动卫星影像视觉问答（VQA）发展至关重要，尤其在气候相关应用中，如灾害监测、基础设施风险评估、城市韧性规划和政策支持。现有VQA模型虽能规模化解析遥感数据，但常缺乏处理复杂地理查询所需的结构化推理能力。本文提出一种融合CoT推理与直接偏好优化（DPO）的VQA框架，通过生成中间推理过程，显著提升检测、分类、空间关系及比较分析等任务的表现，从而为高风险气候领域提供更可靠的决策支持。实验表明，引入CoT监督使准确率较直接基线提升34.9%，而DPO进一步提升了准确率与推理质量。该系统推动了多光谱地球观测下的VQA发展，实现了更丰富的地理空间推理能力，增强了气候应用场景的有效性。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。