3D Morphable Models (MMs) are valuable tools in face perception research. Training of a MM requires vertex-to-vertex correspondence for all face scans in a database. At ZIB, we are developing face matching methods to establish highly accurate dense correspondence in fully automatic fashion. This project aims to improve our current matching pipeline. Specifically, the goal is to utilize geometric deep learning to regress vertex-to-vertex correspondence from face scans directly.

Please see the attachment for further details.

 

 

Research Group: 
Attachements: 
Evaluation and Comparison of Linear and Non-linear Approaches for the Analysis and Synthesis of Facial Motion
MA-01/19