contains material from
Template Matching Techniques in Computer Vision: Theory and Practice
Roberto Brunelli © 2009 John Wiley & Sons, Ltd
When the probability distribution of the templates is unknown, the design of a classifier becomes more complex and many critical estimation issues surfaces. This chapter presents basic results upon which two interrelated, powerful classifier design paradigms stand: regularization networks and support vector machines (SVMs). Several practical hints on how to best use SVM classifiers are described. The techniques are applied to the tasks of gender and race classification based on face images.
keywords: regularization networks, support vector machines, virtual set method, reproducing kernel Hilbert space.