Gender recognition based on face image

被引：1

作者：

Zhang, Meiyan ^{[1
]}

Sun, Jinwei ^{[1
]}

Liu, Dan ^{[1
]}

Wang, Qisong ^{[1
]}

机构：

[1] Harbin Inst Technol, Sch Instrumentat Sci & Engn, Harbin, Peoples R China

来源：

2020 INTERNATIONAL CONFERENCE ON INTELLIGENT COMPUTING AND HUMAN-COMPUTER INTERACTION (ICHCI 2020) | 2020年

基金：

中国国家自然科学基金;

关键词：

component; formatting; style; styling; insert;

D O I：

10.1109/ICHCI51889.2020.00037

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The research on the biometrics recognition of facial image has made great progress. Compared with other biometrics, facial features are stable, simple, intuitive, friendly, non-intrusive, and easily accepted by people. Therefore, facial gender recognition technologies have been successfully applied to many commercial fields. However, the current gender recognition methods still have shortcomings such as low recognition rate and easily affected by surroundings. For this reason, a gender recognition method based on BP neural network is proposed. Firstly, this paper preprocesses the face images, extracts feature of face images, designs a BP neural network and uses the feature parameters to train BP neural network. Afterwards, a classifier based on the face image is obtained. Finally, the classifier is tested using pictures from the database, if demands are not met, BP neural network parameters would be redesigned and the training would be conducted on the BP neural network.

引用

页码：134 / 137

页数：4