arXiv · 2611.16408
Gravity and Flight: Learning to Navigate Fine-Grained Remote Sensing Scenes
Abstract
Vision-language models are increasingly capable, yet their reasoning in complex remote sensing scenes can remain coarse-grained and insufficiently grounded in task-relevant visual evidence. We propose Gravity and Flight (G&F), a framework for scene specialization with limited supervision that separates visual and reasoning verification while coupling policy updates. It combines Evidence-of-Thought supervised fine-tuning (EoT-SFT) and Relational Transfer Policy Optimization (RTPO). EoT-SFT teaches fine-grained observation and faithful reasoning through description, analysis, and supported conclusions. RTPO independently samples responses to scene-matched image pairs with fine-grained visual differences under a shared task. Its rubric-based verification combines Gravity, which checks visual fidelity and description coverage, and Flight, which takes descriptions as premises to assess evidence use and relations between paired judgments. Their scores and task correctness yield a shared pair-level advantage that couples policy updates from both responses. Using 1,008 demonstrations and 500 training pairs for harbor specialization, G&F achieves a mean score of 69.12 across six evaluation sets, exceeding the backbone and supervised initialization by 9.76 and 4.56 points. Matched controls show gains over single-image and joint-input RL, and removing either verification view reduces performance. Applying G&F to urban specialization also improves on the base model and supervised initialization across three evaluation subsets. Project details and updates are available at https://github.com/Z0ngL1ng/Gravity-and-Flight.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Yuheng Zong, Minghua Wang, Xin Zhao, Xiaobin Liu, Yongjian Deng, Chanho Eom. 2026-11-13. Gravity and Flight: Learning to Navigate Fine-Grained Remote Sensing Scenes. https://arxiv.science/abs/2611.16408
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.