Compressed sensing recovery via nonconvex shrinkage penalties

DOI10.1088/0266-5611/32/7/075004zbMATH Open1347.65111arXiv1504.02923OpenAlexW3098595037MaRDI QIDQ3188784FDOQ3188784

Authors: Joseph T. Woodworth, Rick Chartrand 

Publication date: 12 August 2016

Published in: Inverse Problems (Search for Journal in Brave)

Abstract: The $el{l}^0$ minimization of compressed sensing is often relaxed to $el{l}^1$, which yields easy computation using the shrinkage mapping known as soft thresholding, and can be shown to recover the original solution under certain hypotheses. Recent work has derived a general class of shrinkages and associated nonconvex penalties that better approximate the original $el{l}^0$ penalty and empirically can recover the original solution from fewer measurements. We specifically examine p-shrinkage and firm thresholding. In this work, we prove that given data and a measurement matrix from a broad class of matrices, one can choose parameters for these classes of shrinkages to guarantee exact recovery of the sparsest solution. We further prove convergence of the algorithm iterative p-shrinkage (IPS) for solving one such relaxed problem.

Full work available at URL: https://arxiv.org/abs/1504.02923

Recommendations

• Restricted isometry properties and nonconvex compressive sensing
• Sparse recovery by non-convex optimization - instance optimality
• A class of null space conditions for sparse recovery via nonconvex, non-separable minimizations
• Sparse signal recovery via non-convex optimization and overcomplete dictionaries
• An unbiased approach to compressed sensing

zbMATH Keywords

convergencestabilitycompressed sensingrelaxationexact recoverynonconvexityiterative \(p\)-shrinking algorithmparse signals

Mathematics Subject Classification ID

Numerical mathematical programming methods (65K05) Signal theory (characterization, reconstruction, filtering, etc.) (94A12) Nonconvex programming, global optimization (90C26)

Cited In (39)

• Sparse signals recovered by non-convex penalty in quasi-linear systems
• Computational approaches to non-convex, sparsity-inducing multi-penalty regularization
• Compressive Sensing on Manifolds Using a Nonparametric Mixture of Factor Analyzers: Algorithm and Performance Bounds
• Image restoration via the adaptive \(TV^p\) regularization
• Robust nonconvex sparse optimization for impact force identification
• Local recovery bounds for prior support constrained compressed sensing
• Sparsest representations and approximations of an underdetermined linear system
• A singular value \(p\)-shrinkage thresholding algorithm for low rank matrix recovery
• One-bit compressed sensing via \(\ell_p\) \((0<p<1)\)-minimization method
• Non-convex and convex coupling image segmentation via TGpV regularization and thresholding
• Non-uniform recovery guarantees for binary measurements and infinite-dimensional compressed sensing
• Nonconvex sorted \(\ell_1\) minimization for sparse approximation
• Dequantizing Compressed Sensing: When Oversampling and Non-Gaussian Constraints Combine
• Fast L1-L2 minimization via a proximal operator
• Minimization of transformed \(L_1\) penalty: theory, difference of convex function algorithm, and robust application in compressed sensing
• Sparse recovery based on the generalized error function
• An efficient nonconvex regularization for wavelet frame and total variation based image restoration
• Title not available (Why is that?)
• Recent trends and advances in solving the inverse problem for EEG source localization
• Doubly iteratively reweighted algorithm for constrained compressed sensing models
• A first-order smoothed penalty method for compressed sensing
• Limited-angle CT reconstruction with generalized shrinkage operators as regularizers
• Sparse signal recovery based on majorization-minimization with enhanced sparsity
• Compressive Sensing Reconstruction With Prior Information by Iteratively Reweighted Least-Squares
• Two-stage image segmentation based on nonconvex \(\ell_2-\ell_p\) approximation and thresholding
• The springback penalty for robust signal recovery
• Iterative \(p\)-shrinkage thresholding algorithm for low Tucker rank tensor recovery
• A nonconvex nonsmooth image prior based on the hyperbolic tangent function
• Sparse parameter identification for stochastic systems based on \(L_\gamma\) regularization
• Linearized Bregman iterations for compressed sensing
• Minimization of \(L_1\) over \(L_2\) for sparse signal recovery with convergence guarantee
• Restricted isometry properties and nonconvex compressive sensing
• A non-convex regularization approach for compressive sensing
• Parametrized quasi-soft thresholding operator for compressed sensing and matrix completion
• Sampling from non-smooth distributions through Langevin diffusion
• Multiplicative noise removal via using nonconvex regularizers based on total variation and wavelet frame
• Guarantees of fast band restricted thresholding algorithm for low-rank matrix recovery problem
• Linear program relaxation of sparse nonnegative recovery in compressive sensing microarrays
• A class of null space conditions for sparse recovery via nonconvex, non-separable minimizations

Uses Software

• S+WAVELETS

This page was built for publication: Compressed sensing recovery via nonconvex shrinkage penalties

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3188784)