A Novel Face Recognition Method Based on ICA and Binary Tree SVM
Tao Wang · 2017
In this paper, we analyze the fundamental principles, and pros and cons of ICA and SVM, which are commonly used for face recognition. After investigating the SVM-based multiclass classification algorithm, we propose an improved binary tree SVM method, which is then combined with ICA to recognize faces. Features are first extracted via ICA in the experiment on the ORL face dataset. The improved binary tree SVM is then used for recognition. Experimental results demonstrate the effectiveness of the proposed algorithm.