High-Speed Face Recognition Based on Discrete Cosine Transform and RBF Neural Networks

Abstract

In this paper, an efficient method for high-speed face recognition based on the discrete cosine transform (DCT), the Fisher's linear discriminant (FLD) and radial basis function (RBF) neural networks is presented. First, the dimensionality of the original face image is reduced by using the DCT and the large area illumination variations are alleviated by discarding the first few low-frequency DCT coefficients. Next, the truncated DCT coefficient vectors are clustered using the proposed clustering algorithm. This process makes the subsequent FLD more efficient. After implementing the FLD, the most discriminating and invariant facial features are maintained and the training samples are clustered well. As a consequence, further parameter estimation for the RBF neural networks is fulfilled easily which facilitates fast training in the RBF neural networks. Simulation results show that the proposed system achieves excellent performance with high training and recognition speed, high recognition rate as well as very good illumination robustness.

Feature Fusion Using Automatic Generated RBF Neural Network

“…This is also stated in previous studies (Adams and Payandeh, 1996;Chen and Billings, 1992;Er et al, 2005;Gonzalez et al, 2003;Hong et al, 2003;Peng et al, 2006;Li and Wang, 2014;Li et al, 2004Li et al, , 2006 Oyang et al, 2005;Park and Sandberg, 1991;Polycarpou, 2001;Xie and Leung, 2005;Zhang et al, 2004;Zhu and Billings, 1996). Gaussian radial basis functions have also been widely used in support vector machines, an important class of machine learning algorithms.…”

Locating landmarks using templates

“…Here the same steps are used as in the previous paragraph. Er et al (2005) use a training set of 40 grey scale images for each of 10 persons, altogether 400 images. The aim is again the face recognition.…”

“…We would like to mention that Er et al (2005) is an example of many papers in which instead of the whole faces their standardized versions are examined. In this case only the inner parts of faces are used, without hair and without background.…”

Face Recognition System Using PCA and DCT in HMM

“…One of important things for extract the effective features and also for reducing computational complexity in classification stage is Dimensionality reduction.Principal component analysis (PCA) [7], [8], Discrete cosine transform (DCT) [9], and Linear discriminate analysis (LDA) [10] are the main techniques used for data reduction and feature extraction in the appearance based approaches.The most efforts are given mainly on developing feature extraction methods and employing powerful classifiers such as Euclidean distance Classifier, Hidden Markov Models (HMMs) [11], and neural networks [12], [13]. This paper presents a new approach using one dimensional Discrete HMM as classifier and Principal component analysis (PCA) coefficients as features for face recognition afterapplying Discrete cosine transform (DCT).Using seven-states HMM to model face configuration.…”