Acquiring Non Linear Subspace for Face Recognition using Symbolic Kernel PCA Method
نویسنده
چکیده
In this paper, a new technique called symbolic kernel Principal Component Analysis (KPCA) is explored to develop a model for face representation and recognition. The conventional kernel PCA method extracts single valued features from the original image space to represent face images. The proposed method reduces the dimensionality of original image space by representing the face images as symbolic objects (symbolic faces) of interval type variables. Then symbolic kernel PCA method is employed to compute non-linear interval type features from symbolic faces. These features form the non-linear subspace, which maximally preserve original face image information with reduced dimension. A minimum distance classifier with symbolic dissimilarity measure is used for classification. We compare the recognition results of proposed symbolic kernel PCA method with the eigenface method and conventional kernel PCA method using ORL database. Experimental results show that symbolic kernel PCA method with polynomial kernel of degree three achieves improved recognition rate.
منابع مشابه
Contextual Constraints Based Kernel Discriminant Analysis for Face Recognition
In this paper, an improved subspace learning method using contextual constraints based linear discriminant analysis (CCLDA) is proposed for face recognition. The linear CCLDA approach does not consider the higher order nonlinear information in facial images. However, the wide face variations posed by some factors, such as viewpoint, illumination and expression, existing in non-linear subspaces ...
متن کاملFace Recognition Using Kernel Methods
Principal Component Analysis and Fisher Linear Discriminant methods have demonstrated their success in face detection, recognition, and tracking. The representation in these subspace methods is based on second order statistics of the image set, and does not address higher order statistical dependencies such as the relationships among three or more pixels. Recently Higher Order Statistics and In...
متن کاملA comparison of subspace analysis for face recognition
We report the results of a comparative study on subspace analysis methods for face recognition. In particular, we have studied four different subspace representations and their ‘kernelized’ versions if available. They include both unsupervised methods such as Principal Component Analysis (PCA) and Independent Component Analysis (ICA), and supervised methods such as Fisher Discriminant Analysis ...
متن کاملFace Recognition using Eigenfaces , PCA and Supprot Vector Machines
This paper is based on a combination of the principal component analysis (PCA), eigenface and support vector machines. Using N-fold method and with respect to the value of N, any person’s face images are divided into two sections. As a result, vectors of training features and test features are obtain ed. Classification precision and accuracy was examined with three different types of kernel and...
متن کاملRecognizing Faces using Kernel Eigenfaces and Support Vector Machines
In face recognition, Principal Component Analysis (PCA) is often used to extract a low dimensional face representation based on the eigenvector of the face image autocorrelation matrix. Kernel Principal Component Analysis (Kernel PCA) has recently been proposed as a non-linear extension of PCA. While PCA is able to discover and represent linearly embedded manifolds, Kernel PCA can extract low d...
متن کامل