A Generative Variational Model for Inverse Problems in Imaging
From MaRDI portal
Publication:5065477
DOI10.1137/21M1414978zbMATH Open1485.49042arXiv2104.12630MaRDI QIDQ5065477FDOQ5065477
Authors: Andreas Habring, M. Holler 
Publication date: 21 March 2022
Published in: SIAM Journal on Mathematics of Data Science (Search for Journal in Brave)
Abstract: This paper is concerned with the development, analysis and numerical realization of a novel variational model for the regularization of inverse problems in imaging. The proposed model is inspired by the architecture of generative convolutional neural networks; it aims to generate the unknown from variables in a latent space via multi-layer convolutions and non-linear penalties, and penalizes an associated cost. In contrast to conventional neural-network-based approaches, however, the convolution kernels are learned directly from the measured data such that no training is required. The present work provides a mathematical analysis of the proposed model in a function space setting, including proofs for regularity and existence/stability of solutions, and convergence for vanishing noise. Moreover, in a discretized setting, a numerical algorithm for solving various types of inverse problems with the proposed model is derived. Numerical results are provided for applications in inpainting, denoising, deblurring under noise, super-resolution and JPEG decompression with multiple test images.
Full work available at URL: https://arxiv.org/abs/2104.12630
Recommendations
- Variational inference for computational imaging inverse problems
- Deep Generative Models and Inverse Problems
- Solving Bayesian inverse problems from the perspective of deep generative networks
- Bayesian multiscale deep generative model for the solution of high-dimensional inverse problems
- VI-DGP: a variational inference method with deep generative prior for solving high-dimensional inverse problems
- Adversarially learned iterative reconstruction for imaging inverse problems
- Variational models for image inpainting
- Solution of physics-based Bayesian inverse problems with deep generative priors
- Bilevel approaches for learning of variational imaging models
- Variational regularization in inverse problems and machine learning
data sciencemachine learninginverse problemsdeep image priormathematical imaginggenerative neural networks
Artificial neural networks and deep learning (68T07) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Inverse problems in optimal control (49N45) Numerical solutions of ill-posed problems in abstract spaces; regularization (65J20)
Cites Work
- A convex variational model for learning convolutional image atoms from incomplete data
- Nonlinear total variation based noise removal algorithms
- A wavelet tour of signal processing. The sparse way.
- Title not available (Why is that?)
- Secrets of image denoising cuisine
- Total generalized variation
- Variational methods in imaging
- A convergence rates result for Tikhonov regularization in Banach spaces with non-smooth operators
- Lower semicontinuity and relaxation results in BV for integral functionals with BV integrands
- A total variation-based JPEG decompression model
- On representer theorems and convex regularization
- Mathematical image processing. Translated from the German
- Sparsity of solutions for variational inverse problems with finite-dimensional data
- Inertial proximal alternating linearized minimization (iPALM) for nonconvex and nonsmooth problems
- NETT: solving inverse problems with deep neural networks
- Solving inverse problems using data-driven models
- Higher-order total variation approaches and generalisations
- Regularization by architecture: a deep prior approach for inverse problems
- Multi-Layer Convolutional Sparse Modeling: Pursuit and Dictionary Learning
- Deep synthesis network for regularizing inverse problems
- Computed tomography reconstruction using deep image prior and learned reconstruction methods
- A function space framework for structural total variation regularization with applications in inverse problems
- Convolutional Sparse and Low-Rank Coding-Based Image Decomposition
- Sample-Efficient Algorithms for Recovering Structured Signals From Magnitude-Only Measurements
Cited In (18)
- Adversarially learned iterative reconstruction for imaging inverse problems
- Variational networks: an optimal control approach to early stopping variational methods for image restoration
- Constrained and unconstrained deep image prior optimization models with automatic regularization
- A Note on the Regularity of Images Generated by Convolutional Neural Networks
- Deep Generative Models and Inverse Problems
- Generative modeling of convolutional neural networks
- Regularising inverse problems with generative machine learning models
- Learning maximally monotone operators for image recovery
- Blind image inpainting with sparse directional filter dictionaries for lightweight CNNs
- Bilevel inverse problems in neuromorphic imaging
- Multi-layer state evolution under random convolutional design
- Regularization theory of the analytic deep prior approach
- Continuous Generative Neural Networks: A Wavelet-Based Architecture in Function Spaces
- Neural-network-based regularization methods for inverse problems in imaging
- Solving Inverse Problems by Joint Posterior Maximization with Autoencoding Prior
- Robustness and exploration of variational and machine learning approaches to inverse problems: an overview
- Deep unfolding of a proximal interior point method for image restoration
- A convex variational model for learning convolutional image atoms from incomplete data
Uses Software
This page was built for publication: A Generative Variational Model for Inverse Problems in Imaging
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5065477)
