Feature subset selection applied to model-free gait recognition

作者：

Highlights：

• A feature selection framework is proposed to achieve high performance model-free gait recognition.

• The feature selection mechanism relies on the Random Forest algorithm.

• Regions selected are more robust to covariates while reducing the computational cost.

• Panoramic gait recognition is achieved under covariate conditions.

摘要

•A feature selection framework is proposed to achieve high performance model-free gait recognition.•The feature selection mechanism relies on the Random Forest algorithm.•Regions selected are more robust to covariates while reducing the computational cost.•Panoramic gait recognition is achieved under covariate conditions.

论文关键词：Feature selection,Gait recognition,Model-free,Panoramic,Random forest

论文评审过程：Received 3 June 2012, Revised 1 February 2013, Accepted 11 April 2013, Available online 7 May 2013.

论文官网地址：https://doi.org/10.1016/j.imavis.2013.04.001