Enhanced wide-activated residual network for efficient and accurate image deblocking

作者：

Highlights：

• Enhanced wide-activated residual block boosts the information flowability.

• Overlapping patches extraction and combination strategy enlarges the receptive field.

• A novel deep CNN achieves efficient and accurate image deblocking.

摘要

•Enhanced wide-activated residual block boosts the information flowability.•Overlapping patches extraction and combination strategy enlarges the receptive field.•A novel deep CNN achieves efficient and accurate image deblocking.

论文关键词：Image deblocking,Convolutional neural network,Wide activation,Overlapping patches extraction,Overlapping patches combination

论文评审过程：Received 20 January 2020, Revised 2 May 2020, Accepted 6 April 2021, Available online 21 April 2021, Version of Record 24 April 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116283