Insights Into Analysis Operator Learning: From Patch-Based Sparse Models to Higher Order MRFs
From MaRDI portal
(Redirected from Publication:4592059)
Abstract: This paper addresses a new learning algorithm for the recently introduced co-sparse analysis model. First, we give new insights into the co-sparse analysis model by establishing connections to filter-based MRF models, such as the Field of Experts (FoE) model of Roth and Black. For training, we introduce a technique called bi-level optimization to learn the analysis operators. Compared to existing analysis operator learning approaches, our training procedure has the advantage that it is unconstrained with respect to the analysis operator. We investigate the effect of different aspects of the co-sparse analysis model and show that the sparsity promoting function (also called penalty function) is the most important factor in the model. In order to demonstrate the effectiveness of our training approach, we apply our trained models to various classical image restoration problems. Numerical experiments show that our trained models clearly outperform existing analysis operator learning approaches and are on par with state-of-the-art image denoising algorithms. Our approach develops a framework that is intuitive to understand and easy to implement.
Cited in
(41)- Mini-workshop: Deep learning and inverse problems. Abstracts from the mini-workshop held March 4--10, 2018
- A bimodal co-sparse analysis model for image processing
- Fast approximations of shift-variant blur
- Variational networks: an optimal control approach to early stopping variational methods for image restoration
- Field of experts regularized nonlocal low rank matrix approximation for image denoising
- Ultrasound image restoration based on a learned dictionary and a higher-order MRF
- Understanding image inpainting with the help of the Helmholtz equation
- \(\ell^1\)-analysis minimization and generalized (co-)sparsity: when does recovery succeed?
- Adaptive image processing: first order PDE constraint regularizers and a bilevel training scheme
- An abstract convergence framework with application to inertial inexact forward-backward methods
- Structural changes in nonlocal denoising models arising through bi-level parameter learning
- Image denoising via multiscale nonlinear diffusion models
- Poisson Noise Reduction with Higher-Order Natural Image Prior Model
- Sparsity and nullity: paradigms for analysis dictionary learning
- Bilevel optimization with nonsmooth lower level problems
- Techniques for gradient-based bilevel optimization with non-smooth lower level problems
- Discriminative Nonlinear Analysis Operator Learning: When Cosparse Model Meets Image Classification
- Bilevel optimization: theory, algorithms, applications and a bibliography
- A high order PDE-constrained optimization for the image denoising problem
- Convergence of a piggyback-style method for the differentiation of solutions of standard saddle-point problems
- Modern regularization methods for inverse problems
- Solving inverse problems using data-driven models
- An introduction to continuous optimization for imaging
- Bilevel Methods for Image Reconstruction
- Explainable bilevel optimization: an application to the Helsinki Deblur Challenge
- Learning sparsity-promoting regularizers using bilevel optimization
- Learning weakly convex regularizers for convergent image-reconstruction algorithms
- Convergent data-driven regularizations for CT reconstruction
- Robustness and exploration of variational and machine learning approaches to inverse problems: an overview
- Bilevel parameter learning for higher-order total variation regularisation models
- An adaptively inexact method for bilevel learning using primal-dual-style differentiation
- Bilevel learning with inexact stochastic gradients
- An adaptively inexact first-order method for bilevel optimization with application to hyperparameter learning
- Regularization of inverse problems: deep equilibrium models versus bilevel learning
- Linesearch-enhanced forward-backward methods for inexact nonconvex scenarios
- Controlled learning of pointwise nonlinearities in neural-network-like architectures
- Dyadic partition-based training schemes for TV/TGV denoising
- A primal-dual algorithm for image reconstruction with input-convex neural network regularizers
- Product-of-Gaussian-mixture diffusion models for joint nonlinear MRI reconstruction
- Data-driven regularization with weak convexity for robust image reconstruction
- The structure of optimal parameters for image restoration problems
This page was built for publication: Insights Into Analysis Operator Learning: From Patch-Based Sparse Models to Higher Order MRFs
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4592059)
