Browse Books

Reviewer: Adina Raclariu

Over the past decade, computer graphics has exploded out of university research laboratories onto television and movie screens and into medical research laboratories. The authors' goal in writing this book was to survey the state of the art in facial animation and related image synthesis research in order to explain the underlying theory. The book is intended for those interested in pursuing research in facial animation and image systems. It could also be useful in a course on facial animation. After a short history of facial animation, the book enumerates areas in which facial animation can be applied, including computer games, medicine, teleconferencing, social agents, and avatars. The relationship between computer facial animation and conventional animation is outlined. The authors briefly describe the 12 principles of animation, which are universally applied to the production of computer character animation. Next, they present the results of Darwin and Duchenne's research into facial expression analysis and explain their significance. They also cover the Facial Action Coding System developed by Paul Ekman and Wallace Friesen, which breaks down facial action into small units and has been used to help animators interpret and construct realistic facial expressions. The second chapter gives some simplified anatomical terminology and a simplified description of the facial features that are useful for computer synthesis. Essentially, the description comprises two elements: the facial skeleton and the facial muscles. Chapter 3 presents the problems implied by facial modeling, including determining geometric descriptions, animation capabilities, and the colors and textures of the surfaces. The face has a complex, flexible, three-dimensional surface, but facial models that approximate only part of the facial anatomy can be useful. The structure of the model determines its animation potential. In facial animation, the pos sible description techniques are implicit surfaces, parametric surfaces, and polygonal surfaces. The authors discuss various ways in which surfaces can be described and discuss sources of facial surface data. Most of the facial models included rely on measured three-dimensional surface data. The techniques used for surface measurement are digitizers, photogrammetric techniques, and laser scanning techniques. An interesting and useful discussion of those techniques is presented. Chapter 4 depicts the fundamental animation techniques: interpolation, performance-driven, direct-parameterization, pseudomuscle-based, and muscle-based animation. The goal is to manipulate the surfaces of the face over time so that the face has the desired expression in each frame of an animated sequence. Different ways to solve the complex interaction between facial tissue, muscles, and bones, and between the muscles themselves, are depicted. Results in facial control obtained by other researchers are presented and correlated. One of the most important approaches to animation is the use of animation languages. This book discusses the animation programming language developed at Pixar and the Mimic Language developed by Hjortsjo. Some animation examples, which illustrate a few of the concepts discussed in the book, are presented. Chapter 5 describes some of the techniques used to generate the sequence of facial images that form a desired animation. The main tasks involved in image synthesis are covered. Chapter 6 describes parameterized face models, which are motivated by the difficulties of key-pose interpolation. Several possible extensions to direct parameterized models are discussed. The remaining chapters present recent results in facial animation, such as the simulation of skin tissue dynamics, speech synchronization, and the modeling and animation of hair. This book is the most comprehensive one available on facial animation. The problems raised by this complex task are outlined, and the visualization techniques that can be used are analyzed.