167
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
Chapter 10
DOI: 10.4018/978-1-60960-477-6.ch010
INTRODUCTION
In the last couple of decades, automatic comput-
erised face recognition techniques have been de-
veloped for the security of accessing confidential
information in both the internet virtual world and
the real world. Recognition systems are required
to achieve high performance in a variety of dif-
ferent environments such as wide camera angles,
different illuminations and various expressions.
Identity subspace learning techniques have
improved significantly which can be traced
back to Eigen-faces (Turk, 1991) designed for
face recognition. Linear Discriminate Analysis
(LDA) (Belhumeur, 1997), Active Appearance
Models (AAM) (Cootes, 2001), Independent
Hui Fang
Swansea University, UK
Nicolas Costen
Manchester Metropolitan University, UK
Phil Grant
Swansea University, UK
Min Chen
Swansea University, UK
Advances in Moving
Face Recognition
ABSTRACT
This chapter describes the approaches to extracting features via the motion subspace for improving face
recognition from moving face sequences. Although the identity subspace analysis has achieved reasonable
recognition performance in static face images, more recently there has been an interest in motion-based
face recognition. This chapter reviews several state-of-the-art techniques to exploit the motion informa-
tion for recognition and investigates the permuted distinctive motion similarity in the motion subspace.
The motion features extracted from the motion subspaces are used to test the performance based on a
verifcation experimental framework. Through experimental tests, the results show that the correlations
between motion eigen-patterns signifcantly improve the performance of recognition.