Gender classification using face recognition

Abstract

This paper addresses the issue of gender classification using the method of Principal Component Analysis (PCA) for face recognition and classification of human faces. The use of the PCA algorithm has a maximum success rate of 82%. The gender classification system is then improved by using the Linear Discriminant Analysis (LDA. This algorithm has a machine-learning framework by which it trains on a database and using this trained environment to predict the outcome of other images. The classification is restricted to two classes - male and female. Upon using LDA, the success rate increased to approximately 85%. The database used in this paper for the training and testing of images is called the FERET database.