Supervised Manifold Learning and Kernel Independent Component Analysis Applied to the Face Image Recognition

Today the independent component analysis (ICA) has been widely used in the blind source separation (BSS) to separate independent components in a data set based on its statistical properties. However, when the dimension of the input data is too high, the performance of the ICA may be not satisfactory. To address this problem, the present paper has proposed the new integrated method for the independent component extraction. The supervised manifold learning was firstly adopted to reduce the dimension of the input data, and then the kernel ICA (KICA) was employed to extract useful independent components in an efficient manner. The application of the proposed method has been successfully applied to the face image recognition. The experimental analysis has showed satisfactory and effective face image identification performance.