Modelling facial expressions: A framework for simulating nonlinear soft tissue deformations using embedded 3D muscles

Human face can be seen as a soft tissue organ complex with a large investing network of musculature. Due to its complexity, most existing computational models approximate these muscular structures using simple geometries such as 1-D curves or primitive 3-D shapes. This paper presents a new approach to evaluate muscle contribution from anatomically accurate geometries while maintaining the computational complexity at a tractable level. In the proposed method, 3-D muscle structures are embedded inside a facial continuum (encompassing all superficial soft tissue structures), where mechanical contribution of muscles is evaluated independently and transferred to the facial computational domain through a finite element mapping procedure. Muscle forces are decomposed into an array of discrete point loads that are determined at the integration points of an appropriate quadrature scheme. As a result, muscle meshes can be constructed independent from the facial mesh giving two main advantages: (i) the muscle geometries can be refined independent of the facial computational domain, and (ii) it is not required for the computational domain to conform to complex topology of muscle structures.