Pseudo low rank video representation

作者：

Highlights：

• This is the first attempt that exploits eLR AR without any pre-extraction of optical flows, hand-crafted features or demand of high-resolution videos.

• pLRN is the first work that imposes pseudo low rank regularization to networks for getting robust video representation.

• TenneT is an effective data-driven network initialization strategy that can be generalized to other nets.

摘要

•This is the first attempt that exploits eLR AR without any pre-extraction of optical flows, hand-crafted features or demand of high-resolution videos.•pLRN is the first work that imposes pseudo low rank regularization to networks for getting robust video representation.•TenneT is an effective data-driven network initialization strategy that can be generalized to other nets.

论文关键词：Pseudo low rank,Data driven,Low resolution,Action recognition

论文评审过程：Received 26 February 2018, Revised 10 July 2018, Accepted 31 July 2018, Available online 3 August 2018, Version of Record 15 September 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.07.033