Fully transformer network for skin lesion analysis

Medical Image Analysis(2022)

引用 35|浏览29
暂无评分
摘要
•We propose an FTN which is fully relied on Transformer.•We leverage sliding window tokenization to construct hierarchical features.•Spatial Pyramid Transformer to increase efficiency.•Transformer Decoder is proposed to aggregate hierarchical features.
更多
查看译文
关键词
41A05,41A10,65D05,65D17,Skin lesion segmentation,Disease classification,Attention,Transformer,Skin lesion,
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