Simultaneous segmentation and classification of human actions in video streams using deeply optimized Hough transform

作者：

Highlights：

• We offer a learning process for Hough transform.

• This method outperforms other Hough method on honeybee dataset.

• We apply this new method on human action segmentation.

• We evaluate the pipeline on TUM and UTKAD datasets.

• Our results are superior to the best published ones.

摘要

Highlights•We offer a learning process for Hough transform.•This method outperforms other Hough method on honeybee dataset.•We apply this new method on human action segmentation.•We evaluate the pipeline on TUM and UTKAD datasets.•Our results are superior to the best published ones.

论文关键词：Human actions,Segmentation,Classification,Video streams,Hough

论文评审过程：Received 28 October 2013, Revised 10 March 2014, Accepted 12 May 2014, Available online 26 May 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.05.010