目录

目录

Learning Implicit Functions for Topology-Varying Dense 3D Shape Correspondence


Learning implicit functions for topology-varying dense 3D shape correspondence

Motivation

  • 给定一组3D shapes(点云),category-specific model 无监督地学出逐pair(source 与target之间)的correspondence和部件segmentation
  • 即使拓扑不一样也能学到相关性
  • correspondence相关性分数输出
    https://longtimenohack.com/posts/paper_reading/2020nips_liu_learning/image-20201229144020259.png

Overview