arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2609.35227eess.SP

多分辨率块坐标即插即用图像重建算法

Multiresolution Block-Coordinate Plug-and-Play Algorithm for Image Reconstruction

  • ENS de Lyon(里昂高等师范学院)
  • CNRS(法国国家科学研究中心)
  • Inria(法国国家信息与自动化研究所)
  • Université Claude Bernard Lyon 1(里昂第一大学)
  • LIP(信息处理实验室)
  • CPE Lyon(里昂国立高等化学物理电子学院)

机构由 AI 辅助整理,请以论文原文为准。

Edgar Desainte-Mar{é}ville, Marion Foare, Paulo Gon{\c c}alves, Nelly Pustelnik, Elisa Riccietti

AI总结:

针对即插即用图像重建方法难以扩展的问题,提出多分辨率块坐标算法,通过小波分解与随机激活规则选择部分块去噪,在去模糊任务中优于标准方法。

AI中文摘要:

即插即用方法是图像重建领域最先进的方法之一,但其性能随图像尺寸增大而难以扩展。我们提出一种多分辨率块坐标即插即用算法,该算法将图像分解为小波块,并在每次迭代中仅对部分块应用块级去噪网络,这些块通过一种受随机高斯-索斯韦尔启发的激活规则进行选择。去模糊数值结果表明,在不同退化强度下,所提方法优于标准即插即用方法,尤其是自适应规则能动态调整以适应退化强度,在每种情境下均匹配最优激活规则。

英文摘要:

Plug-and-Play methods are among the state-of-the-art approaches for image reconstruction, but they do not scale well with the image size. We propose a multiresolution block-coordinate Plug-and-Play algorithm that decomposes the image into wavelet blocks and applies a block-wise denoiser network to only a subset of blocks at each iteration, selected through a stochastic Gauss-Southwell-inspired activation rule. Numerical results on deblurring show that the proposed method outperforms the standard Plug-and-Play in different degradation regimes, and in particular the adaptive rule dynamically adjusts to the degradation regime to match the optimal activation rule in every context.

↑