Stagewise Training for Hybrid-Distorted Image Restoration
混合失真图像复原的分阶段训练作者机构:School of Information and ElectronicsBeijing Institute of TechnologyBeijing 100081China
出 版 物:《Journal of Shanghai Jiaotong university(Science)》 (上海交通大学学报(英文版))
年 卷 期:2023年第28卷第6期
页 面:793-801页
核心收录:
学科分类:08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程]
主 题:image restoration stagewise training hybrid distortion single distortion
摘 要:Image restoration is the problem of restoring a real degraded *** studies mostly focused on single ***,most of the real images experience multiple distortions,and single distortion image restoration algorithms can not effectively improve the image ***,few existing hybrid distortion image restoration algorithms can not deal with single ***,an end-to-end pipeline network based on stagewise training is proposed in this ***,the network selects three typical image restoration tasks:denoising,inpainting,and super *** whole training process is divided into single distortion training,hybrid distortion training of two types,and hybrid distortion training of three *** design of loss function draws on the idea of deep *** results prove that the proposed method is not only superior to other methods in hybrid-distorted image restoration,but also suitable for single distortion image restoration.