EGAN: Non-uniform image deblurring based on edge adversarial mechanism and partial weight sharing network
作者：
Highlights：
• Dedicate designed edge adversarial mechanism for facilitating the discrimination of image edges.
• A partial weight sharing network benefits for fine image details reconstruction.
• Multiple loss functions drive the network toward the correct manifold.
• Several different blocks learn the high dimensional features representations.
• The method is robust on the synthesized images and real-world images.
摘要
•Dedicate designed edge adversarial mechanism for facilitating the discrimination of image edges.•A partial weight sharing network benefits for fine image details reconstruction.•Multiple loss functions drive the network toward the correct manifold.•Several different blocks learn the high dimensional features representations.•The method is robust on the synthesized images and real-world images.
论文关键词：Generative Adversarial Networks,Image deblurring
论文评审过程：Received 10 August 2019, Revised 15 April 2020, Accepted 1 June 2020, Available online 22 July 2020, Version of Record 5 August 2020.
论文官网地址：https://doi.org/10.1016/j.image.2020.115952
