用等变去噪器提升图像修复的通用性与稳定性
Equivariant Denoisers for Plug and Play Image Restoration
- 设计等变去噪器，让模型对旋转翻转等变换保持一致响应
- 提出两种新框架，在不依赖特定网络结构下实现稳定修复
- 适合需要鲁棒性修复的工业图像处理场景
图像修复的关键在于为干净图像定义合理的先验以补全观测缺失信息。当前先进方法依赖神经网络编码此先验，但典型图像分布对旋转、翻转等变换具有不变性，而多数深度架构无法有效建模这种不变性。近期工作通过在即插即用范式中引入等变性来解决该问题。本文提出两种统一框架：基于去噪的等变正则化（ERED）和等变即插即用（EPnP），均基于等变去噪器与随机优化。我们分析了所提算法的收敛性，并讨论其实际优势。
原文摘要 · Abstract (English)
One key ingredient of image restoration is to define a realistic prior on clean images to complete the missing information in the observation. State-of-the-art restoration methods rely on a neural network to encode this prior. Typical image distributions are invariant to some set of transformations, such as rotations or flips. However, most deep architectures are not designed to represent an invariant image distribution. Recent works have proposed to overcome this difficulty by including equivariance properties within a Plug-and-Play paradigm. In this work, we propose two unified frameworks named Equivariant Regularization by Denoising (ERED) and Equivariant Plug-and-Play (EPnP) based on equivariant denoisers and stochastic optimization. We analyze the convergence of the proposed algorithms and discuss their practical benefit.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。