Gender Classification with GAL Network
Özlem Polat · 2019
This paper presents a solution for gender classification problem using GAL network which is a model of neural networks. This network is trained and tested on a publicly available Stanford dataset including 400 images (200 female and 200 male). Before classification process, primarily Principal Component Analysis (PCA) is applied on 128×128 pixel face images to reduce dimension and the dimension is reduced by about 99%; then GAL, an incremental neural network for supervised learning, is used to classify the data as male or female. Ten-fold cross validation is used for determining the accuracy rate of the classification. This paper has a novelty in the way of applying GAL to face images for gender classification. Test results show that 96% classification performance is obtained with GAL.