Review of deep learning techniques for gender classification in images

Abstract

Automatic gender classification from the face images is a challenging as well as demanding task. It has many applications in the fields of biometrics, security, surveillance, human–computer interaction, etc. Gender recognition requires powerful features of images. Researchers working in this area had proposed different methods of extracting features from image for gender recognition. Some of such features are the Local Binary Patterns (LBP), Scale-Invariant Feature Transform (SIFT), Histogram of Oriented Gradients (HOG), weighted HOG, COSFIRE filter, etc. Beyond this, Convolutional Neural Network (CNN) nowadays are getting widely used for feature extraction and classification in different vision applications. Here, a review is proposed on the use of CNN for gender recognition. The review is supported by results derived through the experiments performed on GENDER-FERET face dataset.