Regularising inverse problems with generative machine learning models

DOI10.1007/S10851-023-01162-XarXiv2107.11191OpenAlexW3183492725MaRDI QIDQ6154435FDOQ6154435

Matthias J. Ehrhardt, Neill D. F. Campbell, M. A. G. Duff

Publication date: 15 February 2024

Published in: Journal of Mathematical Imaging and Vision (Search for Journal in Brave)

Abstract: Deep neural network approaches to inverse imaging problems have produced impressive results in the last few years. In this paper, we consider the use of generative models in a variational regularisation approach to inverse problems. The considered regularisers penalise images that are far from the range of a generative model that has learned to produce images similar to a training dataset. We name this family extit{generative regularisers}. The success of generative regularisers depends on the quality of the generative model and so we propose a set of desired criteria to assess generative models and guide future research. In our numerical experiments, we evaluate three common generative models, autoencoders, variational autoencoders and generative adversarial networks, against our desired criteria. We also test three different generative regularisers on the inverse problems of deblurring, deconvolution, and tomography. We show that restricting solutions of the inverse problem to lie exactly in the range of a generative model can give good results but that allowing small deviations from the range of the generator produces more consistent results.

Full work available at URL: https://arxiv.org/abs/2107.11191

zbMATH Keywords

machine learninginverse problemsimaginggenerative models

Mathematics Subject Classification ID

Computational learning theory (68Q32) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10) Numerical methods for inverse problems for initial value and initial-boundary value problems involving PDEs (65M32)

Cites Work

• Title not available (Why is that?)
• $rm K$-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation
• Nonlinear total variation based noise removal algorithms
• Title not available (Why is that?)
• Title not available (Why is that?)
• Decoding by Linear Programming
• The Little Engine that Could: Regularization by Denoising (RED)
• Proximal alternating linearized minimization for nonconvex and nonsmooth problems
• A first-order primal-dual algorithm for convex problems with applications to imaging
• Title not available (Why is that?)
• Stochastic seismic waveform inversion using generative adversarial networks as a geological prior
• Mathematical image processing. Translated from the German
• A Generative Variational Model for Inverse Problems in Imaging
• NETT: solving inverse problems with deep neural networks
• Solving inverse problems using data-driven models
• Global Guarantees for Enforcing Deep Generative Priors by Empirical Risk
• Modern regularization methods for inverse problems
• Inverse problems. Tikhonov theory and algorithms
• Regularization by architecture: a deep prior approach for inverse problems
• An Introduction to Variational Autoencoders
• Solving Inverse Problems by Joint Posterior Maximization with Autoencoding Prior

Cited In (3)

• Iteratively Refined Image Reconstruction with Learned Attentive Regularizers
• Normalizing flow regularization for photoacoustic tomography
• Neural-network-based regularization methods for inverse problems in imaging

This page was built for publication: Regularising inverse problems with generative machine learning models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6154435)