MLRMV: Multi-layer representation for multi-view action recognition

作者：

Highlights：

• Use the multiview information to reconstruct the daily action recognition approach.

• Design the modified unsupervised clustering algorithm to filter the AP results.

• Present discontinuous temporal scale motion phrases and a phase grading mechanism.

摘要

Highlights•Use the multiview information to reconstruct the daily action recognition approach.•Design the modified unsupervised clustering algorithm to filter the AP results.•Present discontinuous temporal scale motion phrases and a phase grading mechanism.

论文关键词：Multi-layer representation,Multi-view action recognition,Motion atom,Motion phrase

论文评审过程：Received 2 August 2021, Revised 21 October 2021, Accepted 28 October 2021, Available online 3 November 2021, Version of Record 13 November 2021.

论文官网地址：https://doi.org/10.1016/j.imavis.2021.104333