A Novel Feature Extraction Approach to Face Recognition Based on Partial Least Squares Regression
In this paper, partial least square (PLS) regression is firstly employed in image processing. And a new technique coined partial least squares (PLS) regression, line-based PLS, is proposed for feature extraction of the images. To test this new approach, a series of experiments were performed on the famous face image database: ORL face database. Compared with newly proposed two dimensional principal component analysis (2DPCA), it can be found that the dimension of the feature vectors of the line-based PLS is no more than half of the 2DPCA’s while the recognition rate can retain at the same high level. Thus, the feature extraction based on line-based PLS regression is a feasible and effective method.
KeywordsFeature Vector Partial Little Square Face Recognition Recognition Accuracy Latent Vector
Unable to display preview. Download preview PDF.
- 1.Wold, H.: Estimation of Principal Components and Related Model by Iterative Least Squares. In: Multivariate Analysis, pp. 391–420. Academic Press, New York (1966)Google Scholar
- 4.Anthony, R.M., Natasha, R., Nanch, J.L.: Interactions of Prefrontal Cortex in Relation to Awareness in Sensory Learning. Science 284 (1999)Google Scholar
- 8.Huang, D.S.: Systematic Theory of Neural Networks for Pattern Recognition. Publishing House of Electronic Industry of China, Beijing (1996)Google Scholar
- 10.Herve, A.: Partial Least Squares (PLS) Regression. Encyclopedia of Social Sciences Research Method (2003)Google Scholar
- 12.Wang, H.W.: Partial Least-Squares Regression Method and Applications. National Defence Industrial University Press (1999)Google Scholar