Mixtures of Gaussian process models for human pose estimation

作者：

Highlights：

• Novel algorithm for large scale human pose estimation problems.

• Uses multiple Gaussian processes in a mixture of expert framework.

• Allows the accurate regression of Gaussian processes to be scaled to large data.

• Algorithm gives state of the art performance on 3 pose estimation data sets.

摘要

•Novel algorithm for large scale human pose estimation problems.•Uses multiple Gaussian processes in a mixture of expert framework.•Allows the accurate regression of Gaussian processes to be scaled to large data.•Algorithm gives state of the art performance on 3 pose estimation data sets.

论文关键词：Computer vision,Gaussian processes,Human pose estimation,Mixture of experts

论文评审过程：Received 13 January 2013, Revised 26 July 2013, Accepted 23 September 2013, Available online 6 October 2013.

论文官网地址：https://doi.org/10.1016/j.imavis.2013.09.007