Solving ill-posed inverse problems using iterative deep neural networks
From MaRDI portal
Abstract: We propose a partially learned approach for the solution of ill posed inverse problems with not necessarily linear forward operators. The method builds on ideas from classical regularization theory and recent advances in deep learning to perform learning while making use of prior information about the inverse problem encoded in the forward operator, noise model and a regularizing functional. The method results in a gradient-like iterative scheme, where the "gradient" component is learned using a convolutional network that includes the gradients of the data discrepancy and regularizer as input in each iteration. We present results of such a partially learned gradient scheme on a non-linear tomographic inversion problem with simulated data from both the Sheep-Logan phantom as well as a head CT. The outcome is compared against FBP and TV reconstruction and the proposed method provides a 5.4 dB PSNR improvement over the TV reconstruction while being significantly faster, giving reconstructions of 512 x 512 volumes in about 0.4 seconds using a single GPU.
Recommendations
- NETT: solving inverse problems with deep neural networks
- Deep learning for inverse problems with unknown operator
- Big in Japan: regularizing networks for solving inverse problems
- Deep neural networks can stably solve high-dimensional, noisy, non-linear inverse problems
- Learnable Douglas-Rachford iteration and its applications in DOT imaging
Cited in
(only showing first 100 items - show all)- Mini-workshop: Deep learning and inverse problems. Abstracts from the mini-workshop held March 4--10, 2018
- Big in Japan: regularizing networks for solving inverse problems
- Networks for nonlinear diffusion problems in imaging
- A data and knowledge driven approach for SPECT using convolutional neural networks and iterative algorithms
- Model reduction and neural networks for parametric PDEs
- Infinite-dimensional inverse problems with finite measurements
- Estimating adsorption isotherm parameters in chromatography via a virtual injection promoting double feed-forward neural network
- Learning phase field mean curvature flows with neural networks
- Solving optical tomography with deep learning
- Compressive sensing and neural networks from a statistical learning perspective
- Deep learning architectures for nonlinear operator functions and nonlinear inverse problems
- Approximation spaces of deep neural networks
- Solving inverse problems using conditional invertible neural networks
- Determining the defect locations and sizes in elastic plates by using the artificial neural network and boundary element method
- Fast Bayesian inversion for high dimensional inverse problems
- Learning and correcting non-Gaussian model errors
- A Helmholtz equation solver using unsupervised learning: application to transcranial ultrasound
- Deep learning for inverse problems. Abstracts from the workshop held March 7--13, 2021 (hybrid meeting)
- Feasibility-based fixed point networks
- Learning deep implicit Fourier neural operators (IFNOs) with applications to heterogeneous material modeling
- Designing rotationally invariant neural networks from PDEs and variational methods
- Learnable Douglas-Rachford iteration and its applications in DOT imaging
- A review on deep learning in medical image reconstruction
- Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
- A physics-informed variational DeepONet for predicting crack path in quasi-brittle materials
- Nonlocal kernel network (NKN): a stable and resolution-independent deep neural network
- Deep learning for inverse problems with unknown operator
- Linearized inverse Schrödinger potential problem with partial data and its deep neural network inversion
- Data-consistent neural networks for solving nonlinear inverse problems
- Solving traveltime tomography with deep learning
- The first 100 years of the Radon transform
- Deep learning for trivial inverse problems
- Convergence Rates of Spectral Regularization Methods: A Comparison between Ill-Posed Inverse Problems and Statistical Kernel Learning
- scientific article; zbMATH DE number 1329956 (Why is no real title available?)
- Projected regression method for solving Fredholm integral equations arising in the analytic continuation problem of quantum physics
- Deep null space learning for inverse problems: convergence analysis and rates
- Deep convolutional framelets: a general deep learning framework for inverse problems
- Recent trends and advances in solving the inverse problem for EEG source localization
- Equivariant neural networks for inverse problems
- Optimization with learning-informed differential equation constraints and its applications
- Inverse problems on low-dimensional manifolds
- A new approach to solve inverse problems: combination of model-based solving and example-based learning
- Wasserstein-based projections with applications to inverse problems
- Task adapted reconstruction for inverse problems
- An efficient algorithm to compute the X-ray transform
- NETT: solving inverse problems with deep neural networks
- A Data-Driven Iteratively Regularized Landweber Iteration
- Computed tomography reconstruction using deep image prior and learned reconstruction methods
- Numerical solution of inverse problems by weak adversarial networks
- Data driven regularization by projection
- Variational inference for computational imaging inverse problems
- Deep synthesis network for regularizing inverse problems
- Shallow neural networks for fluid flow reconstruction with limited sensors
- Data-driven nonsmooth optimization
- Deep unfolding of a proximal interior point method for image restoration
- Solving inverse problems using data-driven models
- First arrival traveltime tomography using supervised descent learning technique
- Learning the invisible: a hybrid deep learning-shearlet framework for limited angle computed tomography
- On learned operator correction in inverse problems
- Deep neural networks for inverse problems with pseudodifferential operators: an application to limited-angle tomography
- Generalized intersection algorithms with fixed points for image decomposition learning
- Unsupervised knowledge-transfer for learned image reconstruction
- Approximation of discontinuous inverse operators with neural networks
- Deep neural networks can stably solve high-dimensional, noisy, non-linear inverse problems
- Gauss-Newton method for solving linear inverse problems with neural network coders
- Divide-and-conquer DNN approach for the inverse point source problem using a few single frequency measurements
- An introduction to the mathematics of deep learning
- A data-driven Kaczmarz iterative regularization method with non-smooth constraints for ill-posed problems
- Relaxation approach for learning neural network regularizers for a class of identification problems
- In focus -- hybrid deep learning approaches to the HDC2021 challenge
- A direct sampling-based deep learning approach for inverse medium scattering problems
- DRIP: deep regularizers for inverse problems
- Invertible residual networks in the context of regularization theory for linear inverse problems
- A Data-Assisted Two-Stage Method for the Inverse Random Source Problem
- Learning Regularization Parameter-Maps for Variational Image Reconstruction Using Deep Neural Networks and Algorithm Unrolling
- TNet: A Model-Constrained Tikhonov Network Approach for Inverse Problems
- Rethinking the ill-posedness of the spectral function reconstruction -- why is it fundamentally hard and how artificial neural networks can help
- Semi-supervised invertible neural operators for Bayesian inverse problems
- A neural network warm-start approach for the inverse acoustic obstacle scattering problem
- PnP-ReG: Learned Regularizing Gradient for Plug-and-Play Gradient Descent
- On Learning the Invisible in Photoacoustic Tomography with Flat Directionally Sensitive Detector
- Application of machine learning regression models to inverse eigenvalue problems
- Deep unfolding as iterative regularization for imaging inverse problems
- Numerical methods for identifying the diffusion coefficient in a nonlinear elliptic equation
- The mathematics of artificial intelligence
- Solving inverse problems with deep learning
- Reconstruction of inhomogeneous media by an iteration algorithm with a learned projector
- The deep arbitrary polynomial chaos neural network or how deep artificial neural networks could benefit from data-driven homogeneous chaos theory
- Vision graph U-Net: geometric learning enhanced encoder for medical image segmentation and restoration
- Deep unrolling networks with recurrent momentum acceleration for nonlinear inverse problems
- Gabor-filtered Fourier neural operator for solving partial differential equations
- Learning a stable approximation of an existing but unknown inverse mapping: application to the half-time circular Radon transform
- An accelerated inexact Newton regularization scheme with a learned feature-selection rule for non-linear inverse problems
- Training adaptive reconstruction networks for blind inverse problems
- Neural and spectral operator surrogates: unified construction and expression rate bounds
- Sparse regularized CT reconstruction: an optimization perspective
- Learned iterative reconstruction
- Bilevel optimization methods in imaging
- Regularization of inverse problems by neural networks
- Shearlets: from theory to deep learning
This page was built for publication: Solving ill-posed inverse problems using iterative deep neural networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4601414)
