Combining K-Means Attention and Hierarchical Mimicking Strategy for 3D U-Net Based Brain Tumor Segmentation

2021 IEEE International Conference on Information Communication and Software Engineering (ICICSE)(2021)

引用 2|浏览0
暂无评分
摘要
Despite the progress of U-Net in medical image segmentation, standard U-Net still has the problem of generalization ability and limited receptive fields. This paper proposes an improved 3D U-Net which focuses on model generalization and receptive fields for the semantic segmentation of brain tumor. Inspired by the success of using Hierarchical Mimicking (HM) in instance recognition, we present a n...
更多
查看译文
关键词
3D UNet,segmentation,hierarchical mimicking,K-means attention
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要