A novel face recognition method based on Principal Component Analysis and Kernel Partial Least Squares
Xiujuan Li, Jie Ma, Shutao Li · 2007
The automatic recognition of human face presents a significant challenge to the pattern recognition research community recently. As a traditional method for face recognition, principal component analysis (PCA) may neglect differentia. Since kernel partial least squares (KPLS) creates orthogonal score vectors by using the existing correlations and keeps most of the variance between different classes, it can compensate for the defects of PCA. This paper proposes a novel feature extraction method using PCA and KPLS and classification using support vector machine (SVM). The experimental results based on ORL and Yale face database prove that the proposed method has excellent classification performance.