Publication year: 2006
Internet Resource: Please Login to download book
This volume provides a comprehensive view of unconstrained face recognition, especially face recognition from multiple still images and/or video sequences, assembling a collection of novel approaches able to recognize human faces under various unconstrained situations. The underlying basis of these approaches is that, unlike conventional face recognition algorithms, they exploit the inherent characteristics of the unconstrained situation and thus improve the recognition performance when compared with conventional algorithms. Unconstrained Face Recognition is accessible to a wide audience with an elementary level of linear algebra, probability and statistics, and signal processing.