Facial expression cloning with fuzzy set clustering

Abstract

This paper describes work towards a facial expression cloning system that maps pre-existing generic facial expressions from a shape-based face to a muscle-based face. The goal of this work is to create a facial expression cloning algorithm that allows animations to be recorded for one face and transferred to other faces regardless of the underlying model used for the source and target faces. Facial expressions are compared using membership values of fuzzy sets.