Perceived Emotionality of Linear and Non-Linear AUs Synthesised using a 3D Dynamic Morphable Facial Model

Research using dynamic facial expressions in computer science and psychology is largely focused on facial models with control parameters based on the Facial Action Coding System (FACS) [Ekman et al. 2002]. Facial models used in research and production are typically linear in nature, whereas real expressions are non-linear. Using a 3D Dynamic Morphable Model [Cosker et al. 2010], in this work we explore the effect of linear and non-linear facial movement on expression recognition. We believe that this has implications in the validity of using linear or non-linear models in facial experiments, and also impacts on the design of facial models in general.