cs.CVNov 12, 2026

Gravity and Flight: Learning to Navigate Fine-Grained Remote Sensing Scenes

Authors: Yuheng Zong, Minghua Wang, Xin Zhao, Xiaobin Liu, Yongjian Deng, Chanho Eom

Organizations: The Institute of Robotics and Automatic Information System (IRAIS), the Tianjin Key Laboratory of Intelligent Robotics (tjKLIR), Nankai University, Tianjin, 300071, China · The College of Artificial Intelligence, Nankai University, Tianjin, 300071, China · The Tianjin Key Laboratory of Intelligent Robotics, the Academy for Advanced Interdisciplinary Studies, and the College of Artificial Intelligence, Nankai University, Tianjin, 300071, China · The Department of Imaging Science and Arts & Metaverse Convergence, Chung-Ang University, South Korea

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.

Figures & tables

Appendix figures & tables8 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Benchmarking Fine-Grained Visual Reasoning for Remote Sensing in Urban Management

    Nov 4, 2026Jiawei Zhang, Rizhao Cai, Dayan GuanRemote Sensing Image UnderstandingReasoning for VLMs

  2. GeoSolver: Scaling Test-Time Reasoning in Remote Sensing with Fine-Grained Process Supervision

    Mar 10, 2026Lang Sun, Ronghao Fu, Zhuoran Duan +3VLM ReasoningProcess Reward Models

  3. RS-HyRe-R1: A Hybrid Reward Mechanism to Overcome Perceptual Inertia for Remote Sensing Images Understanding

    Apr 19, 2026Gaozhi Zhou, Hu He, Peng Shen +8Vision-Language ModelsVisual Question Answering