A framework for joint estimation of age, gender and ethnicity on a large database

作者：

Highlights：

• A framework for joint estimation of age, gender and ethnicity in a single step;

• A novel finding on feature dimensionality in estimating age, gender and ethnicity;

• A rank theory based analysis of dimensionality problem in using CCA based methods;

• A ranking of CCA and PLS based methods under our joint estimation framework;

• Investigation of LS formulations of the CCA based methods for our problem.

摘要

•A framework for joint estimation of age, gender and ethnicity in a single step;•A novel finding on feature dimensionality in estimating age, gender and ethnicity;•A rank theory based analysis of dimensionality problem in using CCA based methods;•A ranking of CCA and PLS based methods under our joint estimation framework;•Investigation of LS formulations of the CCA based methods for our problem.

论文关键词：Joint estimation of age, gender and ethnicity,A general framework,Partial least squares (PLS),Canonical correlation analysis (CCA),Regularized CCA,Kernel PLS (KPLS),Kernel CCA (KCCA),Regularized KCCA,Least squares CCA

论文评审过程：Received 31 May 2013, Revised 10 April 2014, Accepted 30 April 2014, Available online 10 May 2014.

论文官网地址：https://doi.org/10.1016/j.imavis.2014.04.011