CORE-Seg: Reasoning-Driven Segmentation for Complex Lesions via Reinforcement Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917317768642560 |
|---|---|
| author | Xie, Yuxin Chen, Yuming Yang, Yishan Zhou, Yi Zhou, Tao Zhao, Zhen Liu, Jiacheng Fu, Huazhu |
| author_facet | Xie, Yuxin Chen, Yuming Yang, Yishan Zhou, Yi Zhou, Tao Zhao, Zhen Liu, Jiacheng Fu, Huazhu |
| contents | Medical image segmentation is undergoing a paradigm shift from conventional visual pattern matching to cognitive reasoning analysis. Although Multimodal Large Language Models (MLLMs) have shown promise in integrating linguistic and visual knowledge, significant gaps remain: existing general MLLMs possess broad common sense but lack the specialized visual reasoning required for complex lesions, whereas traditional segmentation models excel at pixel-level segmentation but lack logical interpretability. In this paper, we introduce ComLesion-14K, the first diverse Chain-of-Thought (CoT) benchmark for reasoning-driven complex lesion segmentation. To accomplish this task, we propose CORE-Seg, an end-to-end framework integrating reasoning with segmentation through a Semantic-Guided Prompt Adapter. We design a progressive training strategy from SFT to GRPO, equipped with an adaptive dual-granularity reward mechanism to mitigate reward sparsity. Our Method achieves state-of-the-art results with a mean Dice of 37.06\% (14.89\% higher than the second-best baseline), while reducing the failure rate to 18.42\%. Project Page: https://xyxl024.github.io/CORE-Seg.github.io/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_05911 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CORE-Seg: Reasoning-Driven Segmentation for Complex Lesions via Reinforcement Learning Xie, Yuxin Chen, Yuming Yang, Yishan Zhou, Yi Zhou, Tao Zhao, Zhen Liu, Jiacheng Fu, Huazhu Computer Vision and Pattern Recognition Artificial Intelligence Medical image segmentation is undergoing a paradigm shift from conventional visual pattern matching to cognitive reasoning analysis. Although Multimodal Large Language Models (MLLMs) have shown promise in integrating linguistic and visual knowledge, significant gaps remain: existing general MLLMs possess broad common sense but lack the specialized visual reasoning required for complex lesions, whereas traditional segmentation models excel at pixel-level segmentation but lack logical interpretability. In this paper, we introduce ComLesion-14K, the first diverse Chain-of-Thought (CoT) benchmark for reasoning-driven complex lesion segmentation. To accomplish this task, we propose CORE-Seg, an end-to-end framework integrating reasoning with segmentation through a Semantic-Guided Prompt Adapter. We design a progressive training strategy from SFT to GRPO, equipped with an adaptive dual-granularity reward mechanism to mitigate reward sparsity. Our Method achieves state-of-the-art results with a mean Dice of 37.06\% (14.89\% higher than the second-best baseline), while reducing the failure rate to 18.42\%. Project Page: https://xyxl024.github.io/CORE-Seg.github.io/ |
| title | CORE-Seg: Reasoning-Driven Segmentation for Complex Lesions via Reinforcement Learning |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2603.05911 |