Low resolution face recognition using a two-branch deep convolutional neural network architecture

作者：

Highlights：

• A novel nonlinear coupled mapping architecture using two deep convolutional neural networks.

• Achieves high recognition accuracy especially when the probe image is extremely low resolution.

• With the embedded super-resolution CNN, it reconstructs high resolution version of the input image.

摘要

•A novel nonlinear coupled mapping architecture using two deep convolutional neural networks.•Achieves high recognition accuracy especially when the probe image is extremely low resolution.•With the embedded super-resolution CNN, it reconstructs high resolution version of the input image.

论文关键词：Low resolution face recognition,Super-resolution methods,Coupled mappings methods,Deep convolutional neural networks

论文评审过程：Received 4 February 2019, Revised 26 July 2019, Accepted 28 July 2019, Available online 30 July 2019, Version of Record 7 August 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.112854