Cephalometric landmark detection via global and local encoders and patch-wise attentions

Neurocomputing(2022)

引用 4|浏览13
暂无评分
摘要
•Multiple cephalometric landmarks can be localized via a single-passing CNN.•A novel patch-wise attention mechanism is introduced to fuse global and local features.•Multiple cephalometric landmarks can be successfully localized without additional post-processing or the modeling of an anatomical topology.•Variations on head postures and anatomical structures can be handled.
更多
查看译文
关键词
Auxiliary decoder networks,Cephalometric landmark detection,Direct point regression,Explicit global and local feature fusion,Patch-wise 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