Face recognition is one of the most familiar applications of image analysis and has gained much attention in recent years. Several computational methods are implemented in this field, appearance-based subspace analysis still gives the most promising results. This project is a comparative study of the most popular appearance-based face recognition projection methods (PCA and ICA). The three architectures discussed in "Bartlett, M.S., Movellan, J.R., Sejnowski, T.J.: Face Recognition by Independent Component Analysis, IEEE Trans. on Neural Networks, 13(6), (2002) 1450–1464" are implemented in this project, and built- in functions of artificial neural network, and K nearest-neighbors are used as the classification method. Also two popular databases for face recognition (Yale and ORL) are used here in order to test the accuracy of the implemented methods.
The qualitative results of ORL database. (left) The first twenty-five eigenfaces, (middle) the first twenty-five statistically independent basis images from first architecture, (right) the first twenty-five statistically independent basis images from second architecture (ICA factorial representation).
Sid Ali Rezetane, Sina M. Baharlou, Harold Agudelo
Adviser: Prof. Aurelio Uncini
Download Yale and ORL face datasets; set the parameters in nnTest.m and KnnTest.m and run them with matlab.