An effective deep network using target vector update modules for image restoration

作者：

Highlights：

• Effective target vector update modules are proposed for image restoration (IR).

• An effective IR engine network is proposed to drive the framework.

• The IR process is implemented in the transformation domain.

• Experimental results demonstrate the superiority of the proposed method.

摘要

•Effective target vector update modules are proposed for image restoration (IR).•An effective IR engine network is proposed to drive the framework.•The IR process is implemented in the transformation domain.•Experimental results demonstrate the superiority of the proposed method.

论文关键词：Image restoration,Plug and play method,Convolutional neural network framework,Transformation domain,Target vector update module

论文评审过程：Received 18 June 2020, Revised 24 July 2021, Accepted 17 September 2021, Available online 20 September 2021, Version of Record 24 September 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108333