Just how and why we find some faces memorable but others impossible to recognize is a complex and important question with many practical applications. This edited collection brings together a multi-disciplinary team of researchers to compare and contrast their findings from psychology, computer science, mathematics and neural network research to present a state-of-the-art description of the latest and best of their results in this field. With an international line-up of key researches, this book is an resource for all researchers and advanced students of this topic in each discipline.
An initial effort to construct a theoretical base for how people and computers recognize faces, drawing solely from the statistics of the information available in the population of faces. Mathematicians, computer scientists, and neural network theorists share the principle that we do not recognize faces in isolation, and that only by knowing a large number of faces can we judge a particular one to be, for example, female, attractive, oriental, happy, or distinctive. They also query how caricatures are effective. Annotation c. by Book News, Inc., Portland, Or.
