Sparsity based methods for overparameterized variational problems

DOI10.1137/140998585zbMATH Open1328.47080arXiv1405.4969OpenAlexW2962781485MaRDI QIDQ3454487FDOQ3454487

Authors: Raja Giryes, Michael Elad, Alfred M. Bruckstein 

Publication date: 25 November 2015

Published in: SIAM Journal on Imaging Sciences (Search for Journal in Brave)

Abstract: Two complementary approaches have been extensively used in signal and image processing leading to novel results, the sparse representation methodology and the variational strategy. Recently, a new sparsity based model has been proposed, the cosparse analysis framework, which may potentially help in bridging sparse approximation based methods to the traditional total-variation minimization. Based on this, we introduce a sparsity based framework for solving overparameterized variational problems. The latter has been used to improve the estimation of optical flow and also for general denoising of signals and images. However, the recovery of the space varying parameters involved was not adequately addressed by traditional variational methods. We first demonstrate the efficiency of the new framework for one dimensional signals in recovering a piecewise linear and polynomial function. Then, we illustrate how the new technique can be used for denoising and segmentation of images.

Full work available at URL: https://arxiv.org/abs/1405.4969

Recommendations

• Over-parameterized variational optical flow
• A variational formulation for frame-based inverse problems
• On globally optimal local modeling: from moving least squares to over-parametrization
• Sparse template-based variational image segmentation
• Sparsity of solutions for variational inverse problems with finite-dimensional data

zbMATH Keywords

total variationdenoisingsparsityimage segmentationcompressed sensingvariational methodsoverparameterizationcosparsityline segmentation

Mathematics Subject Classification ID

Signal theory (characterization, reconstruction, filtering, etc.) (94A12) Computing methodologies for image processing (68U10) Variational methods applied to PDEs (35A15) Applications of operator theory in optimization, convex analysis, mathematical programming, economics (47N10) Inverse problems in optimal control (49N45) Numerical aspects of computer graphics, image analysis, and computational geometry (65D18) Numerical solution to inverse problems in abstract spaces (65J22) Method of lines for initial value and initial-boundary value problems involving PDEs (65M20)

Cites Work

• Nonlinear total variation based noise removal algorithms
• Title not available (Why is that?)
• Atomic Decomposition by Basis Pursuit
• Model Selection and Estimation in Regression with Grouped Variables
• Matching pursuits with time-frequency dictionaries
• Image Processing and Analysis
• From Sparse Solutions of Systems of Equations to Sparse Modeling of Signals and Images
• Approximation to Data by Splines with Free Knots
• Iterative hard thresholding for compressed sensing
• CoSaMP: Iterative signal recovery from incomplete and inaccurate samples
• Algorithms for simultaneous sparse approximation. II: Convex relaxation
• Geodesic active contours
• Stable image reconstruction using total variation minimization
• Optimal approximations by piecewise smooth functions and associated variational problems
• Approximation of functional depending on jumps by elliptic functional via t-convergence
• Recovery Algorithms for Vector-Valued Data with Joint Sparsity Constraints
• Title not available (Why is that?)
• Sparse solutions to linear inverse problems with multiple measurement vectors
• Adaptive greedy approximations
• Variational methods in image segmentation with 7 image processing experiments
• Sparse representations in unions of bases
• Active contours without edges
• Block-Sparse Signals: Uncertainty Relations and Efficient Recovery
• Algorithms for simultaneous sparse approximation. I: Greedy pursuit
• The cosparse analysis model and algorithms
• Performance Guarantees of the Thresholding Algorithm for the Cosparse Analysis Model
• A Survey on Variational Optic Flow Methods for Small Displacements
• On the Reconstruction of Block-Sparse Signals With an Optimal Number of Measurements
• Analysis versus synthesis in signal priors
• Signal Space CoSaMP for Sparse Recovery With Redundant Dictionaries
• Sparse components of images and optimal atomic decompositions
• Efficient graph-based image segmentation
• Subspace Pursuit for Compressive Sensing Signal Reconstruction
• Model-Based Compressive Sensing
• Compressed sensing with coherent and redundant dictionaries
• On the stability of the basis pursuit in the presence of noise
• Reduce and Boost: Recovering Arbitrary Sets of Jointly Sparse Vectors
• Hard thresholding pursuit: an algorithm for compressive sensing
• Robust Recovery of Signals From a Structured Union of Subspaces
• An Empirical Bayesian Strategy for Solving the Simultaneous Sparse Approximation Problem
• Modern statistical estimation via oracle inequalities
• Over-parameterized variational optical flow
• Orthogonal least squares methods and their application to non-linear system identification
• A Theory for Sampling Signals From a Union of Subspaces
• Sampling Theorems for Signals From the Union of Finite-Dimensional Linear Subspaces
• Near oracle performance and block analysis of signal space greedy methods
• Greedy-like algorithms for the cosparse analysis model
• RIP-Based Near-Oracle Performance Guarantees for SP, CoSaMP, and IHT
• Greedy signal space methods for incoherence and beyond
• On globally optimal local modeling: from moving least squares to over-parametrization

Cited In (6)

• Sparsity-Inducing Nonconvex Nonseparable Regularization for Convex Image Processing
• Sparse approximation based on a random overcomplete basis
• Near oracle performance and block analysis of signal space greedy methods
• Overparameterized Models for Vector Fields
• On globally optimal local modeling: from moving least squares to over-parametrization
• Hyper-sparsity in the revised simplex method and how to exploit it

Uses Software

• PDCO
• CoSaMP

This page was built for publication: Sparsity based methods for overparameterized variational problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3454487)