Principal directions for local independent components analysis
NN'08 Proceedings of the 9th WSEAS International Conference on Neural Networks
An Improved Algorithm for Estimating the ICA Model Concerning the Convergence Rate
AMDO '08 Proceedings of the 5th international conference on Articulated Motion and Deformable Objects
A Version of the FastICA Algorithm Based on the Secant Method Combined with Simple Iterations Method
ICISP '08 Proceedings of the 3rd international conference on Image and Signal Processing
Hi-index | 0.00 |
This paper proposes a face recognition using both a preprocessing method by centroid shift and a fixedpoint (FP) independent component analysis (ICA) based on secant method. The centroid shift is applied to find a validity image by removing the background of source images. FP-ICA of secant method is an alternate of the conventional FP-ICA based on Newton's method, which is to exclude the complex computation of differential process in Newton's method. The proposed method has been applied to a problem which recognizes the 48 Yale face images. The experimental results show that the proposed FPICA method is superior to the conventional in the recognition performance for the test images and that the case preprocessed with the proposed method has better outcome in face recognition rate and processing time than the cases not preprocessed.