Identify computer generated characters by analysing facial expressions variation

Abstract

Significant improvements have been recently achieved in both quality and realism of computer generated characters, which are nowadays often very difficult to be distinguished from real ones. However, generating highly realistic facial expressions is still a challenging issue, since synthetic expressions usually follow a repetitive pattern, while in natural faces the same expression is usually produced in similar but not equal ways. In this paper, we propose a method to distinguish between computer generated and natural faces based on facial expressions analysis. In particular, small variations of the facial shape models corresponding to the same expression are used as evidence of synthetic characters.