Abstract

Wavelet networks (WNs) were introduced in 1992 as a combination of artificial neural RBF networks and wavelet decomposition. Since then, however, WNs have received only little attention. We believe, that the potential of WNs is generally underestimated. WNs have the advantage, that the wavelet coefficients are directly related to the image information through the wavelet transform. In addition, the parameters of the wavelets in the WNs are subject to optimization, which results in a direct relation between the represented function and the optimized wavelets, lead to a considerable data reduction (thus making subsequent algorithms much more efficient) as well as in wavelets that can be used as an optimized filter bank. We have therefore analyzed some of their properties and hightlight their advantages for object representation purposes. We present a series of experimental results where we have used WNs for face tracking in which we exploit the efficiency due to the data reduction, for face recognition and face-pose estimation where we exploit the optimized filter bank principle of the WNs.