Face Representation and Recognition Using Partial Least Squares Regression

The partial least squares(PLS) regression is a new multivariate data analysis method developed from practical applications in real word.In this paper,a new face representation and the recognition method based on PLS model is presented.In contrast to principal component analysis(PCA),the process of extracting lower-dimensional face representation features by PLS contains the class information of samples and has better classification performance.The experimental results on ORL face image database have shown that when the testing images are reconstructed,PLS-based methods are superior to PCA-based ones,and the former classification results is better than the latter.