Gender classification by principal component analysis and support vector machine
Proceedings of the 2011 International Conference on Communication, Computing & Security
Hi-index | 0.00 |
In this paper, an Enhanced PCA-SIFT is proposed and a FSVM is adopted for gender classification. The Enhanced PCA-SIFT is based on PCA-SIFT, which has been successfully applied into feature extraction, the Enhanced PCA-SIFT is to extract face features including gender information. A membership algorithm based on LVQ is used in FSVM. In FERET, CAS-PEAL and BUAA-IRIP face image database, Experimental results prove that the gender classification method proposed in this paper could result in an identification of high accuracy and stability.