Human activity recognition using dynamic representation and matching of skeleton feature sequences from RGB-D images

作者：

Highlights：

•

• This paper studies human activity recognition based on the segmentation technique.

• We cast activity recognition as a dynamic representation and matching problem.

• Frame sets with different numbers of poses are obtained by a learning strategy.

• The shapeDTW algorithm is utilized to measure the distance between feature segments.

摘要

•This paper studies human activity recognition based on the segmentation technique.•We cast activity recognition as a dynamic representation and matching problem.•Frame sets with different numbers of poses are obtained by a learning strategy.•The shapeDTW algorithm is utilized to measure the distance between feature segments.

论文关键词：Human activity recognition,Dynamic representation and matching,Shape dynamic time warping

论文评审过程：Received 18 January 2018, Revised 24 June 2018, Accepted 24 June 2018, Available online 3 July 2018, Version of Record 18 September 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.06.013