In this dissertation, I describe a virtual head model with anatomical structure. The model is animated in a physics-based manner by use of muscle contractions that in turn cause skin deformations; the simulation is efficient enough to achieve real-time frame rates on current PC hardware. Construction of head models is eased in my approach by deriving new models from a prototype, employing a deformation method that reshapes the complete virtual head structure. Without additional modelling tasks, this results in a model that can be immediately animated. The general deformation method allows for several applications, such as adaptation to individual scan data for creation of animated head models of real people. The basis for the deformation method is a set of facial feature points, which leads to other interesting uses when this set is chosen according to an anthropometric standard set of facial landmarks: I present algorithms for the simulation of human head growth and reconstruction of a face from a skull.

The creation of computer-animated human faces is a long-standing and challenging problem since the early 1970s. There are numerous approaches to facial animation, but to this day no general-purpose system exists that solves the problem in a manner satisfying the needs of all practical applications. In the medical field, highly accurate reproduction of a real head is required to enable well-informed decisions in surgery planning. Animation capabilities and computation time are not important. On the other hand, realism is of minor concern in interactive dialog systems or computer games. Here, the animation merely has to look plausible but play in real-time. A real-life model does not need to be reproduced with all subtleties of facial shape and texture.

An obvious approach to achieve generality is the simulation of the inner workings of a real face. High expectations are tied to physics-based systems, where the ultimate goal is to have the full range of conformation and expressiveness in the face emerge "naturally" through precise modeling of the anatomical structure and accurate simulation of tissue properties. This has so far only been realized in parts, and no current implementation catches all the intricacies of the human face.

Traditionally, the computational cost of physics-based simulation has been prohibitive for real-time facial animation on consumer-class PC hardware. This has changed dramatically in recent years, making fast, high-quality animation possible on current desktop computers. But, apart from the run-time issues, constructing a virtual head model with the complex structure of skull, muscles, and skin is non-trivial, requiring artistic skills and time.

This situation motivates my dissertation: I propose that an anatomy-based virtual head model can be animated in real time using numerical simulation techniques, driven by an advanced facial muscle model. The simulation is efficient enough to achieve real-time frame rates on current PC hardware. Manual construction of such a structured head model is a difficult task, and avoided in my approach by deriving new models from a prototype, employing a deformation method that reshapes the complete head structure. Without additional modeling tasks, this results in a model that can be immediately animated. The general deformation method allows for several applications including the adaptation of a reference head model to individual scan data to produce animated head models of real people. The methods and techniques described are demonstrated on human head models, but they are also applicable to other virtual creatures, with few modifications in the construction process.

The deformation method is based on facial feature points, which leads to other interesting uses when an anthropometric standard set of facial landmarks is chosen: using a database of facial measurements, shape attributes of the face can be changed in a controlled, meaningful manner. As an example, I show how statistically plausible growth of an individual head from childhood to adult age can be simulated. Furthermore, based on known skull / skin distance relationships, a face can be approximated from the skull geometry. These applications demonstrate the usefulness of the approach outside the realm of computer graphics, for instance, in the forensic sciences.

In summary, the key contributions made to the field in this dissertation are:

• an anatomy-based head model: the human head's major structural components are built explicitly into its virtual counterpart: a flexible skin surface, an advanced model for virtual muscles controlling the animation, an embedded skull to which skin and muscles attach, as well as separately modeled geometric objects for eyes, teeth, and tongue. A mass-spring system connects skin, muscles, and skull. Additionally, a set of landmarks, defined on the skin and skull surfaces, is part of the model. The model and associated algorithms enable real-time, physics-based animation.
• a general deformation method for the head model: all parts of the virtual head structure are reshaped simultaneously using a landmark-based approach; the resulting model is immediately animatable with no further modeling steps.
• creation of animatable head models from range scan data: an algorithm is presented that employs the deformation method for adaptation of a generic head template to scan data.
• simulation of human head growth: making use of landmark-based anthropometric measurements, the shape changes of the human head from infancy to adulthood are simulated.
• reconstruction of faces from skull data: I propose a technique that allows fast and flexible modeling of face reconstructions from a scanned skull. This has applications mainly in the forensic sciences.

"A Head Model with Anatomical Structure for Facial Modeling and Animation" by Kolja Kaehler, Master's Thesis, 168 Pages, Acrobat PDF