An Effective Method for Gender Classification with Convolutional Neural Networks

摘要

A gender classification system uses a given image from human face to tell the gender of the given person. An effective gender classification approach is able to improve the performance of many other applications, including image or video retrieval, security monitoring, human-computer interaction and so on. In this paper, an effective method for gender classification task in frontal facial images based on convolutional neural networks (CNNs) is presented. Our experiments have been shown that the method of CNNs for gender classification task is effective and achieves higher classification accuracy than others on FERET and CAS-PEAL-R1 facial datasets. Finally, we built a gender classification demo, where input is the scene image per frame captured by the camera and the output is the original scene image with marked on detected facial areas.