Gender classification from facial images

Abstract

Gender is one striking feature that human can deduce effortlessly when looking at a face. Here, we try to classify the gender (male or female) based on the face images. The first part of this paper presents a review of different methods/approaches used for gender recognition. We present a comparative analysis for gender recognition using PCA, 2dPCA and its variants. Finally, we develop an iterative model using 2dPCA which updates itself when new samples are encountered. This model is expected to be fruitful in real-life situation as it can learn when it comes across new test samples. We consider CFD, CUHK, ORL and Yale facial data-sets for our experiments.