{"ID":5554333,"CreatedAt":"2026-07-02T02:11:27.934456424Z","UpdatedAt":"2026-07-04T17:22:49.717376202Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.01191","arxiv_id":"2607.01191","title":"Perceive-to-Reason: Decoupling Perception and Reasoning for Fine-Grained Visual Reasoning","abstract":"Fine-grained visual reasoning remains challenging for vision-language models, especially when small but critical visual cues are buried in high-resolution images. Existing approaches rely on repeated cropping or test-time visual search to introduce local evidence, but they typically do not explicitly distinguish perception from reasoning. In this paper, we propose Perceive-to-Reason (P2R), a unified framework that formulates fine-grained visual reasoning as a two-stage process: the model first localizes question-relevant evidence as a Perceiver, and then answers the question as a Reasoner based on the annotated image and cropped regions. To better align training with this decoupled formulation, we further introduce Perception-Reasoning Alternating GRPO (PRA-GRPO), a role-aware reinforcement learning strategy that alternates between perception-focused and reasoning-focused updates using only final-answer supervision. Built on top of Qwen3-VL-Instruct-2B/4B/8B, P2R consistently improves performance across model scales. In particular, P2R-4B achieves 93.2% on V-Star, 81.9% on HR-Bench-4K, and 80.5% on HR-Bench-8K, substantially outperforming its corresponding backbone. Further experiments show that the benefits of P2R extend beyond high-resolution benchmarks to broader multimodal reasoning tasks. These results suggest that explicitly decoupling perception from reasoning provides an effective framework for fine-grained visual reasoning.","short_abstract":"Fine-grained visual reasoning remains challenging for vision-language models, especially when small but critical visual cues are buried in high-resolution images. Existing approaches rely on repeated cropping or test-time visual search to introduce local evidence, but they typically do not explicitly distinguish percep...","url_abs":"https://arxiv.org/abs/2607.01191","url_pdf":"https://arxiv.org/pdf/2607.01191v1","authors":"[\"Hongxing Li\",\"Xiufeng Huang\",\"Dingming Li\",\"Wenjing Jiang\",\"Zixuan Wang\",\"Haolei Xu\",\"Hanrong Zhang\",\"Haiwen Hong\",\"Longtao Huang\",\"Hui Xue\",\"Weiming Lu\",\"Jun Xiao\",\"Yueting Zhuang\",\"Yongliang Shen\"]","published":"2026-07-01T17:24:26Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Reinforcement Learning\",\"Language Model\"]","has_code":false}