Multiscale parallel deep CNN (mpdCNN) architecture for the real low-resolution face recognition for surveillance

作者：

Highlights：

• The paper proposes a parallel deep CNN architecture called mpdCNN

• mdCNN architecture performs face recognition in real low-resolution images.

• The mpdCNN architecture is a multi-scale architecture.

• The mpdCNN architecture gives 88.6% accuracy on SCface database.

摘要

Highlights•The paper proposes a parallel deep CNN architecture called mpdCNN•mdCNN architecture performs face recognition in real low-resolution images.•The mpdCNN architecture is a multi-scale architecture.•The mpdCNN architecture gives 88.6% accuracy on SCface database.

论文关键词：Face recognition,Low-resolution images,Surveillance,Multiscale parallel architecture,Convolution neural network

论文评审过程：Received 31 March 2021, Revised 7 July 2021, Accepted 3 September 2021, Available online 10 September 2021, Version of Record 28 September 2021.

论文官网地址：https://doi.org/10.1016/j.imavis.2021.104290