Stability and super-resolution of generalized spike recovery

DOI10.1016/J.ACHA.2016.09.004MaRDI QIDQ723008zbMATH OpenOpenAlexFDO

Authors Dmitry Batenkov

Publication date 30 July 2018

Published in Applied and Computational Harmonic Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1409.3137

zbMATH Keywords

super-resolutiondecimationnumerical conditioningprony systemspike recovery

Mathematics Subject Classification ID

Image processing (compression, reconstruction, etc.) in information and communication theory (94A08)

Abstract: We consider the problem of recovering a linear combination of Dirac delta functions and derivatives from a finite number of Fourier samples corrupted by noise. This is a generalized version of the well-known spike recovery problem, which is receiving much attention recently. We analyze the numerical conditioning of this problem in two different settings depending on the order of magnitude of the quantity $Neta$, where $N$ is the number of Fourier samples and $eta$ is the minimal distance between the generalized spikes. In the "well-conditioned" regime $Netagg1$, we provide upper bounds for first-order perturbation of the solution to the corresponding least-squares problem. In the near-colliding, or "super-resolution" regime $Netao0$ with a single cluster, we propose a natural regularization scheme based on decimating the samples extendash{} essentially increasing the separation $eta$ extendash{} and demonstrate the effectiveness and near-optimality of this scheme in practice.

Recommendations

• A theory of super-resolution from short-time Fourier transform measurements
• Super-resolution of near-colliding point sources
• Towards a Mathematical Theory of Super‐resolution
• Super-resolution from noisy data
• Super-resolution of point sources via convex programming

Cites work

• scientific article; zbMATH DE number 432498 (Why is no real title available?)
• scientific article; zbMATH DE number 733519 (Why is no real title available?)
• A Generalization of Hermite's Interpolation Formula
• A Stable Numerical Method for Inverting Shape from Moments
• A generalized Prony method for reconstruction of sparse sums of eigenfunctions of linear operators
• Accurate solution of near-colliding Prony systems via decimation and homotopy continuation
• An ``algebraic reconstruction of piecewise-smooth functions from integral measurements
• Cleaning-up data for sparse model synthesis: when symbolic-numeric computation meets error-correcting codes
• Complete algebraic reconstruction of piecewise-smooth functions from Fourier data
• CramÉr–Rao Bounds for Multiple Poles and Coefficients of Quasi-Polynomials in Colored Noise
• Exponential data fitting and its applications
• Geometry and singularities of the Prony mapping
• Linear versus non-linear acquisition of step-functions
• Local and global geometry of Prony systems and Fourier reconstruction of piecewise-smooth functions
• MUSIC for single-snapshot spectral estimation: stability and super-resolution
• Near Minimax Line Spectral Estimation
• Nonlinear approximation by sums of exponentials and translates
• Numerical Polynomial Algebra
• On inverses of Vandermonde and confluent Vandermonde matrices
• On inverses of Vandermonde and confluent Vandermonde matrices. II
• On the Accuracy and Resolution of Powersum-Based Sampling Methods
• On the accuracy of solving confluent Prony systems
• On the numerical condition of a generalized Hankel eigenvalue problem
• On the problem of parameter estimation in exponential sums
• Parameter estimation for exponential sums by approximate prony method
• Performance Bounds and Design Criteria for Estimating Finite Rate of Innovation Signals
• Performance of ESPRIT for Estimating Mixtures of Complex Exponentials Modulated by Polynomials
• Practical Extrapolation Methods
• Prony systems via decimation and homotopy continuation
• Reconstructing planar domains from their moments
• Sampling Moments and Reconstructing Signals of Finite Rate of Innovation: Shannon Meets Strang–Fix
• Sampling and reconstruction of signals with finite rate of innovation in the presence of noise
• Shape from Moments—An Estimation Theory Perspective
• Signal Acquisition from Measurements via Non-Linear Models
• Signal decomposition and analysis via extraction of frequencies
• Spike detection from inaccurate samplings
• Super-resolution from noisy data
• Super-resolution, extremal functions and the condition number of Vandermonde matrices
• Superresolution via Sparsity Constraints
• Towards a Mathematical Theory of Super‐resolution
• Variable projection for nonlinear least squares problems

Cited in

(14)

• Accuracy of noisy spike-train reconstruction: a singularity theory point of view
• Conditioning of partial nonuniform Fourier matrices with clustered nodes
• SVD update methods for large matrices and applications
• On the accuracy of Prony's method for recovery of exponential sums with closely spaced exponents
• The spectral properties of Vandermonde matrices with clustered nodes
• Stable soft extrapolation of entire functions
• Multiscale matrix pencils for separable reconstruction problems
• Super-resolution of generalized spikes and spectra of confluent Vandermonde matrices
• Super-resolution meets machine learning: approximation of measures
• A diffusion + wavelet-window method for recovery of super-resolution point-masses with application to single-molecule microscopy and beyond
• On algebraic properties of low rank approximations of Prony systems
• A mathematical theory of the computational resolution limit in one dimension
• A note on spike localization for line spectrum estimation
• Super-resolution wavelets for recovery of arbitrarily close point-masses with arbitrarily small coefficients

This page was built for publication: Stability and super-resolution of generalized spike recovery

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q723008)