Machine Recognition in Complex Domain - High Dimensional Neurocomputing: Growth, Appraisal and Applications - page 154

Fig. 7.8 Eigenfaces computed from Indian face database

Fig. 7.9 ICA basis images computed from Indian face database

yielded by ICA-based methods provide faster learning in all neural classifiers. It is
worthwhile to mention that this face dataset has much variations in poses, which
decrease its recognition accuracy. But, performance of C ICA in feature extraction is
significant in terms of precision in the classification and accuracy. With C ICA ,clas-
sification error is quiet low for class faces while it is quiet high for nonclass faces.