团队简介    研究论文    研究人员    发明专利    研究资助    联系我们    English
Biharmonic Deformation Transfer with Automatic Key Point Selection
Accepted to CVM, 2018.

Jie Yang, Lin Gao, Yu-Kun Lai, Paul L. Rosin, Shihong Xia

Abstract
Deformation transfer is an important research problem in geometry processing and computer animation, which greatly simplifies the effort for generating deformation sequences of a target shape by reusing existing deformation of a source shape. A fundamental problem for existing deformation transfer methods is to build reliable correspondences. This is challenging, especially when the source and target shapes differ significantly in appearance, and in practice it is typically achieved with manual labeling. In this paper, we propose a novel deformation transfer method that aims at minimizing user effort. We adapt a biharmonic weight deformation framework which is able to produce plausible deformation even with only a few key points. We then develop an automatic algorithm to identify a minimum set of key points on the source model that characterizes the deformation well. Once the user specifies correspondences for these identified key points on the target model, the deformation is effectively transferred to the target shape using biharmonic deformation. While minimal user effort is still needed, our approach avoids the difficult problem of choosing key points which requires expertise and often involves trial-and-error to obtain desired results. Experimental results demonstrate that our method requiring little user effort produces better deformation results than alternative solutions.
Downloads

Biharmonic Deformation Transfer with Automatic Key Point Selection.  Accepted to CVM, 2018.