CR-Seg: Attention-Guided and CoT-Enhanced Coarse-to-Refined Reasoning Segmentation
CR-Seg:注意力引导与CoT增强的由粗到精推理分割
机构 * School of Computer Science and Engineering, Northeastern University(东北大学计算机科学与工程学院)
专题命中 视觉空间推理 :CoT(title,title_cn);reasoning(title,abstract);chain-of-thought(abstract);分类 cs.AI
AI总结 提出CR-Seg两阶段框架,通过注意力图提取和全局到局部思维链,实现由粗到精的推理分割,解决跨模态对齐和推理-答案不一致问题。