数码设计Issue(24)：106-108,3.
融合归一化流不确定性核估计的去运动模糊算法
A De-Motion Fuzzy Algorithm Incorporating Uncertainty Kernel Estimation for Normalized Flow
王翔立 1余婷婷 1苏章 1汪耀 1朱厚帅1
作者信息
- 1. 武汉轻工大学电气与电子工程学院,武汉 430023
- 折叠
摘要
Abstract
The main work of the article contained two aspects.On the one hand,in order to improve the model's ability to recover image details and obtain clearer images,a fuzzy kernel estimation was introduced into the network,and in order to overcome the problem of the lack of basic facts of non-uniform fuzzy kernel,the fuzzy kernel estimation module was trained in a self-supervised way.On the other hand,a composite loss function was proposed to improve the original generator loss function,and the slope loss was introduced as the discriminator loss function,so as to obtain a higher and more stable training efficiency for the network.关键词
图像处理/模糊算法/图像信息/图像恢复Key words
image processing/fuzzy algorithm/image information/image recovery分类
信息技术与安全科学引用本文复制引用
王翔立,余婷婷,苏章,汪耀,朱厚帅..融合归一化流不确定性核估计的去运动模糊算法[J].数码设计,2024,(24):106-108,3.
