Colmenarez A.J., Xiong Z., Huang T.S. Facial Analysis from Continuous Video with Applications to Human-Computer Interface

Издательство Kluwer, 2004, -159 pp.Automatic facial analysis is one of the most important problems in computer vision and applications of video analysis. Face recognition, though being only one part of facial analysis, has applications in areas such as access control, face databases, face ID card, human computer interaction, law enforcement, multimedia management, security, smart cards and surveillance.
The material in this book is the result of about 10 years of the authors’ research on face detection, facial features detection, face recognition, face tracking, emotion recognition and 3-D face modelling. What is unique about it is its strong information theoretic foundation upon which the underlying algorithms are developed. Because of this strong foundation, the face detection algorithm using the Information-Based Maximum Discrimination criterion remains to be one of the fastest, robust face detection algorithms. In this book we give a detailed description of these algorithms.
We also present a real time system for facial analysis for human-computer interface. Its characteristics are summarized as follows. First, the system is capable of detecting the presence of people in natural scenes with complex backgrounds. Then, it tracks the faces and their features in real time to produce information about the position, facial activity, and history of presence. Next, it analyzes the absolute position of the face in order to detect gestures such as shaking and nodding the head and moving the face forward or away from the camera. Finally, it recognizes faces and facial expressions based on a novel probabilistic framework in which faces are modelled by their appearances and their facial expressions. The book is designed for sharing our research results with graduate students, researchers in facial analysis and computer vision. We look forward to inquiries, collaborations, criticisms and suggestions from them.Information-Based Maximum Discrimination
Face and Facial Feature Detection
Face and Facial Feature Tracking
Face and Facial Expression Recognition
3-D Model-Based Image Communication
Implementations, Experiments and Results
Application in an Audio-Visual Person Recognition System