AI RESEARCH
High-Resolution Visual Reasoning via Multi-Turn Grounding-Based Reinforcement Learning
arXiv CS.CV
•
ArXi:2507.05920v2 Announce Type: replace State-of-the-art large multi-modal models (LMMs) face challenges when processing high-resolution images, as these inputs are converted into enormous visual tokens, many of which are irrelevant to the downstream task. In this paper, we propose Multi-turn Grounding-based Policy Optimization (MGPO), an end-to-end reinforcement learning (RL) framework that enables LMMs to iteratively focus on key visual regions by automatically cropping sub-images, based on model-predicted grounding coordinates within a multi-turn conversation framework.