Wavelet energy entropy as a new feature extractor for face recognition

In face recognition, it is important how to select the invariant facial features especially faces with various pose and expression changes. This paper presents wavelet energy entropy as new facial features to recognize faces under various pose and expression changes. Preliminaiy experiment results on ORL, YALE face databases with different pose and expression changes indicate that the proposed wavelet energy entropy feature face recognition method is fast and effective compared to other tradition algorithms.