Sparse composition of body poses and atomic actions for human activity recognition in RGB-D videos

作者：

Highlights：

• A novel hierarchical model to recognize human activities using RGB-D data is proposed.

• The method jointly learns suitable representations at different abstraction levels.

• The model achieves multi-class discrimination providing useful mid-level annotations.

• The compositional capabilities of our model also bring robustness to body occlusions.

摘要

•A novel hierarchical model to recognize human activities using RGB-D data is proposed.•The method jointly learns suitable representations at different abstraction levels.•The model achieves multi-class discrimination providing useful mid-level annotations.•The compositional capabilities of our model also bring robustness to body occlusions.

论文关键词：Activity recognition,Hierarchical recognition model,RGB-D videos

论文评审过程：Received 26 May 2016, Accepted 16 November 2016, Available online 27 November 2016, Version of Record 10 January 2017.

论文官网地址：https://doi.org/10.1016/j.imavis.2016.11.004