基于递归正交最小二乘的径向基函数网络人脸识别

本文提出了一种基于递归正交最小二乘的径向基函数(RBF)网络人脸识别算法,该算法首先使用主成分分析(PCA)提取输入图像特征,将提取的特征作为RBF网络的输入进行识别,在求取网络权值时采用递归正交最小二乘(ROLS)算法。实验表明,该算法能明显地缩短训练时间同时具有较高的识别率。

Radial Basis Function (RBF) Neural Networks For Face Recognition with Recursive Orthogonal Least Squares

In this paper,we presents an improved algorithm for face recognition with Radial Basis Function (RBF) Neural Net- works.First,a set of features of input images are generated by the Principal Component Analysis (PCA),then the features are used as the inputs of RBF neural networks,Recursive Orthogonal Least Squares(ROLS) is implemented to train the RBF neural networks.Simulation results conducted on the ORL database and Stirling database show that the proposed algorithm achieves excellent performance both in terms of error rates of classification and learning efficiency.