计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (24): 16-25.DOI: 10.3778/j.issn.1002-8331.2305-0110

• 热点与综述 • 上一篇    下一篇

基于判别模型的年龄不变人脸识别方法综述

杨晓艳,邓淼磊,张德贤,李磊,王翠   

  1. 1.河南工业大学 信息科学与工程学院,郑州 450001
    2.河南省粮食信息处理国际联合实验室,郑州 450001
  • 出版日期:2023-12-15 发布日期:2023-12-15

Review of Age-Invariant Face Recognition Methods Based on Discriminant Models

YANG Xiaoyan, DENG Miaolei, ZHANG Dexian, LI Lei, WANG Cui   

  1. 1.College of Information Science and Engineering, Henan University of Technology, Zhengzhou 450001, China
    2.Henan International Joint Laboratory of Grain Information Processing, Zhengzhou 450001, China
  • Online:2023-12-15 Published:2023-12-15

摘要: 人脸识别是一种利用人体面部特征进行身份验证的生物识别技术。但随着年龄的增长,人的面部轮廓以及纹理都会发生很大变化,从而给人脸识别带来了巨大挑战。因此,年龄不变人脸识别(age-invariant face recognition,AIFR)研究具有重要意义。介绍了判别方法的研究现状,包括传统判别方法以及基于深度学习的判别方法,并对优缺点进行梳理总结。梳理了年龄不变人脸识别技术领域内代表性数据集以及常用的评价指标,并将优秀算法的性能在常用数据集上进行了实验比较。对年龄不变人脸识别技术的发展趋势进行了展望。

关键词: 年龄不变人脸识别, 判别方法, 深度学习

Abstract: Face recognition is a biometric technology that uses human facial features for identity verification. But with the increase of age, the contour, shape and texture of the face will change significantly, which brings great challenge to face recognition. Thus, the study of age-invariant face recognition is of great significance. Firstly, this paper introduces the research status of discriminant methods, including traditional discriminant methods and discriminant methods based on deep learning, and summarizes the advantages and disadvantages. Then, the representative datasets and common evaluation indexes in the field of age-invariant face recognition are sorted out, and the performance of excellent algorithms is compared on common datasets. Finally, the future development of age-invariant face recognition technology is discussed.

Key words: age-invariant face recognition, discriminative method, deep learning