计算机科学 ›› 2018, Vol. 45 ›› Issue (6): 284-290.doi: 10.11896/j.issn.1002-137X.2018.06.050

所属专题: 人脸识别

• 图形图像与模式识别 • 上一篇    下一篇

基于标签敏感最大间隔准则的人脸年龄两步估计算法

徐晓玲1,2, 金忠1,2, 贲圣兰3   

  1. 南京理工大学计算机科学与工程学院 南京2100941;
    南京理工大学高维信息智能感知与系统教育部重点实验室 南京2100942;
    南京大学电子科学与工程学院 南京2100233
  • 收稿日期:2017-04-18 出版日期:2018-06-15 发布日期:2018-07-24
  • 作者简介:徐晓玲(1993-),女,硕士生,CCF会员,主要研究领域为计算机视觉、人脸图像分析,E-mail:xuxiaoling1028.njust@hotmail.com;金 忠(1961-),男,博士,教授,主要研究领域为模式识别、图像分析、机器学习、计算机视觉,E-mail:zhongjin@njust.edu.cn(通信作者);贲圣兰(1982-),女,博士,主要研究领域为模式识别、人脸识别
  • 基金资助:
    本文受国家自然科学基金(61373063,61375007,61233011),国家重点基础研究发展计划(2014CB349303)资助

Facial Age Two-steps Estimation Algorithm Based on Label-sensitive Maximum Margin Criterion

XU Xiao-ling1,2, JIN Zhong1,2, BEN Sheng-lan3   

  1. School of Computer Science and Engineering,Nanjing University of Science & Technology,Nanjing 210094,China1;
    Key Laboratory of Intelligent Perception and System for High-Dimensional Information of Ministry of Education,Nanjing University of Science & Technology,Nanjing 210094,China2;
    School of Electronic Science and Engineering,Nanjing University,Nanjing 210023,China3
  • Received:2017-04-18 Online:2018-06-15 Published:2018-07-24

摘要: 传统的最大间隔准则在计算类间离散度矩阵时往往忽略了类别之间的差异,但是对于人脸年龄估计,不同年龄标签之间的差异性是非常显著的。因此,在标签之间引入距离度量,提出标签敏感的最大间隔准则维数约减算法。此外,考虑到人脸变老的复杂性,提出两步的局部回归算法——K近邻-标签分布的支持向量回归(K Nearset Neighbors-Label Distribution Support Vector Reressor,KNN-LDSVR),以进行人脸年龄估计。在FGNET数据库子集上提出的人脸年龄估计方法的平均绝对误差为4.1岁,相对于已有的年龄估计方法,性能得到提升。

关键词: 标签敏感, 局部回归, 两步, 年龄估计, 最大间隔准则

Abstract: Traditional maximum margin criterion usually ignores the differences between classes in the computation of the between-class scatter matrix.However,for facial age estimation,the differences between age labels are very significant.Therefore,this paper proposed a novel dimensionality reduction algorithm,called label-sensitive maximum margin criterion (lsMMC),by introducing a distance metric between the classes.In addition,considering the complicated facial aging process,this paper proposed a two-steps local regression algorithm named K nearest neighbors-label distribution support vector regressor (KNN-LDSVR) for age estimation.The mean absolute error of the proposed facial aging estimation method on the FGNET database subset is 4.1 years,which improves the performance compared with existing age estimation methods.

Key words: Age estimation, Label-sensitive, Local regression, Maximum margin criterion, Two-steps

中图分类号: 

  • TP391.4
[1]GUO G,MU G,FU Y,et al.Human age estimation using bio-inspired features[C]//IEEE International Conference on Compu-ter Vision and Pattern Recognition.2009:112-119.
[2]COOTES T F,EDWARDS G J,TAYLOR C J.Active appearance models[C]//European Conference on Computer Vision.1998:484-498.
[3]DONG Y,LIU Y,LIAN S.Automatic age estimation based on deep learning algorithm[J].Neurocomputing,2016,187:4-10.
[4]MA B,SHAN S,CHEN X,et al.Discriminant analysis for perceptionally comparable classes[C]//IEEE International Confe-rence on Automatic Face & Gesture Recognition.2008:1-6.
[5]CHAO W L,LIU J Z,DING J J.Facial age estimation based on label-sensitive learning and age-oriented regression[J].Pattern Recognition,2013,46(3):628-641.
[6]ONIFADE O F W,AKINYEMI D J.A review on the suitability of machine learning approaches to facial age estimation[J].International Journal of Modern Education and Computer Science,2015,7(12):17-28.
[7]PONTES J K,BRITTO A S,FOOKES C,et al.A flexible hierarchical approach for facial age estimation based on multiple features[J].Pattern Recognition,2016,54(C):34-51.
[8]GENG X,YIN C,ZHOU Z H.Facial age estimation by learning from label distributions[J].IEEE Transactions on Pattern Ana-lysis and Machine Intelligence,2013,35(10):2401-2412.
[9]HE Z,LI X,ZHANG Z,et al.Data-dependent label distribution learning for age estimation[J].IEEE Transactions on Image Processing,2017,26(8):3846-3858.
[10]The FG-NET Aging Database [OL].http://sting.cycollege.ac.cy/~lanitis/fgnetaging.
[11]LI H,JIANG T,ZHANG K.Efficient and robust feature extraction by maximum margin criterion[J].IEEE Transactions on Neural Networks,2006,17(1):157-165.
[12]MIKA S,RATSCH G,WESTON J,et al.Fisher discriminant analysis with kernels[C]//IEEE International Conference on Signal Processing.1999:41-48.
[13]SUGIYAMA M.Dimensionality reduction of multimodal labeled data by local fisher discriminant analysis[J].Journal of Machine Learning Research,2007,8(5):1027-1061.
[14]GENG X,HOU P.Pre-release prediction of crowd opinion on movies by label distribution learning[C]//International Joint Conference on Artificial Intelligence.2015:3511-3517.
[15]GOODALL C.Procrustes methods in the statistical analysis of shape[J].Journal of the Royal Statistical Society,1991,53(2):285-339.
[16]LANITIS A,TAYLOR C J,COOTES T F.Toward automatic simulation of aging effects on face images[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(4):442-455.
[17]GENG X,ZHOU Z H,ZHANG Y,et al.Learning from facial aging patterns for automatic age estimation[C]//ACM International Conference on Multimedia.2006:307-316.
[18]GENG X,ZHOU Z H,SMITH-MILES K.Automatic age estimation based on facial aging patterns[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(12):2234-2240.
[19]CHANG K Y,CHEN C S,HUNG Y P.A ranking approach for human ages estimation based on face images[C]//IEEE International Conference on Pattern Recognition.2010:3396-3399.
[20]CHOI S E,LEE Y J,LEE S J,et al.Age estimation using a hierarchical classifier based on global and local facial features[J].Pattern Recognition,2011,44(6):1262-1281.
[21]WU T,TURAGA P,CHELLAPPA R.Age estimation and face verification across aging using landmarks[J].IEEE Transactions on Information Forensics and Security,2012,7(6):1780-1788.
[22]LIU J,MA Y,DUAN L,et al.Hybrid constraint SVR for facial age estimation[J].Signal Processing,2014,94(1):576-582.
[23]SI J,FENG J,BU Q,et al.Age estimation based on canonical correlation analysis and extreme learning machine[C]//Chinese Conference on Biometric Recognition.2015:677-685.
[24]TIAN Q,XUE H,QIAO L.Human age estimation by considering both the ordinality and similarity of ages[J].Neural Processing Letters,2016,43(2):1-17.
[1] 杜亮, 任鑫, 张海莹, 周芃.
基于局部回归融合的多核聚类方法
Multiple Kernel Clustering via Local Regression Integration
计算机科学, 2021, 48(8): 47-52. https://doi.org/10.11896/jsjkx.201000106
[2] 孙劲光, 荣文钊.
基于区域的年龄估计模型研究
Research on Regional Age Estimation Model
计算机科学, 2018, 45(8): 41-49. https://doi.org/10.11896/j.issn.1002-137X.2018.08.008
[3] 赵阳阳,刘纪平,杨毅,张福浩,仇阿根.
混合时空地理加权回归及参数的两步估计
Mixed Geographically and Temporally Weighted Regression and Two-step Estimation
计算机科学, 2017, 44(3): 274-277. https://doi.org/10.11896/j.issn.1002-137X.2017.03.056
[4] 林时苗,毛晓蛟,杨育彬.
基于人脸图像的年龄估计
Age Estimation Based on Facial Image
计算机科学, 2015, 42(6): 32-36. https://doi.org/10.11896/j.issn.1002-137X.2015.06.007
[5] .
一种移动人体轮廓的两步提取法

计算机科学, 2008, 35(8): 220-222.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!