arXiv 2603.281202026-03-30cs.CV

MedLoc-R1: 为基于GRPO的医疗视觉接地任务的性能感知课程奖励调度

MedLoc-R1: Performance-Aware Curriculum Reward Scheduling for GRPO-Based Medical Visual Grounding

Beijing University of Posts and Telecommunications（北京邮电大学）；Emory University（埃默里大学）；Peking University（北京大学）；Shandong University（山东大学）

Guangjing Yang, Ziyuan Qin, Chaoran Zhang, Chenlin Du, Jinlin Wang, Wanran Sun, Zhenyu Zhang, Bing Ji, Qicheng Lao

备注：: 2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出MedLoc-R1框架，通过性能感知奖励调度提升医疗视觉接地任务的定位准确性和训练稳定性，无需引入辅助网络。

摘要：

医疗视觉接地是细粒度多模态推理和可解释临床决策支持的关键基础。尽管强化学习（RL）在接地任务中取得进展，现有方法如组相对策略优化（GRPO）在直接应用于医学图像时面临严重的奖励稀疏性问题，主要是由于局部化小或模糊区域的固有难度，以及RL中固定IoU奖励方案的刚性和次优性质。为此，我们提出MedLoc-R1，一种性能感知的奖励调度框架，根据模型准备程度逐步收紧奖励标准。MedLoc-R1引入滑动窗口性能跟踪器和多条件更新规则，自动调整奖励调度，从密集易获得信号转向严格精细的局部化要求，同时保持GRPO的有利特性。在三个医疗视觉接地基准测试中，MedLoc-R1在GRPO基线上一致提高了定位准确性和训练稳定性。我们的框架为高风险医疗应用中的基于RL的接地任务提供了一种通用、轻量级和有效的解决方案。代码和检查点可在https://github.com/MembrAI/MedLoc-R1获取。

英文摘要：

Medical visual grounding serves as a crucial foundation for fine-grained multimodal reasoning and interpretable clinical decision support. Despite recent advances in reinforcement learning (RL) for grounding tasks, existing approaches such as Group Relative Policy Optimization~(GRPO) suffer from severe reward sparsity when directly applied to medical images, primarily due to the inherent difficulty of localizing small or ambiguous regions of interest, which is further exacerbated by the rigid and suboptimal nature of fixed IoU-based reward schemes in RL. This leads to vanishing policy gradients and stagnated optimization, particularly during early training. To address this challenge, we propose MedLoc-R1, a performance-aware reward scheduling framework that progressively tightens the reward criterion in accordance with model readiness. MedLoc-R1 introduces a sliding-window performance tracker and a multi-condition update rule that automatically adjust the reward schedule from dense, easily obtainable signals to stricter, fine-grained localization requirements, while preserving the favorable properties of GRPO without introducing auxiliary networks or additional gradient paths. Experiments on three medical visual grounding benchmarks demonstrate that MedLoc-R1 consistently improves both localization accuracy and training stability over GRPO-based baselines. Our framework offers a general, lightweight, and effective solution for RL-based grounding in high-stakes medical applications. Code \& checkpoints are available at \hyperlink{}{https://github.com/MembrAI/MedLoc-R1}.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。