Joint gender classification and age estimation by nearly orthogonalizing their semantic spaces

作者：

Highlights：

• Formulate the semantic relationship between human gender and age as near-orthogonality regularization

• Propose a joint estimation framework for human gender and age based on the semantic regularization

• Exemplify the proposed framework

• Kernelize the proposed framework by deriving a representer theorem

• Experimentally demonstrate the effectiveness and superiority of the proposed methods

摘要

•Formulate the semantic relationship between human gender and age as near-orthogonality regularization•Propose a joint estimation framework for human gender and age based on the semantic regularization•Exemplify the proposed framework•Kernelize the proposed framework by deriving a representer theorem•Experimentally demonstrate the effectiveness and superiority of the proposed methods

论文关键词：Gender classification,Age estimation,Nearly orthogonal semantic spaces,Support vector ordinal regression,Discriminant learning for ordinal regression

论文评审过程：Received 21 October 2016, Revised 26 September 2017, Accepted 31 October 2017, Available online 5 November 2017, Version of Record 20 November 2017.

论文官网地址：https://doi.org/10.1016/j.imavis.2017.10.003