A theory of super-resolution from short-time Fourier transform measurements

DOI10.1007/S00041-017-9534-XMaRDI QIDQ1704865zbMATH OpenOpenAlexFDO

Authors Céline Aubel, David Stotz, Helmut Bölcskei

Publication date 13 March 2018

Published in The Journal of Fourier Analysis and Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1509.01047

zbMATH Keywords

sparsitysuper-resolutioninverse problems in measure spaces

Mathematics Subject Classification ID

Spaces of measures (46E27) Spaces of measures, convergence of measures (28A33) Fourier and Fourier-Stieltjes transforms and other transforms of Fourier type (42B10) Applications of functional analysis in optimization, convex analysis, mathematical programming, economics (46N10) Holomorphic functions of several complex variables (32A10) Topological linear spaces of test functions, distributions and ultradistributions (46F05)

Abstract: While spike trains are obviously not band-limited, the theory of super-resolution tells us that perfect recovery of unknown spike locations and weights from low-pass Fourier transform measurements is possible provided that the minimum spacing, $Delta$, between spikes is not too small. Specifically, for a measurement cutoff frequency of $f_c$, Donoho [2] showed that exact recovery is possible if the spikes (on $mathbbR$) lie on a lattice and $Delta>1/f_c$, but does not specify a corresponding recovery method. Cand$$s and Fernandez-Granda [3, 4] provide a convex programming method for the recovery of periodic spike trains (i.e., spike trains on the torus $mathbbT$), which succeeds provably if $Delta>2/f_c$ and $f_{c}geq128$ or if $Delta>1.26/f_c$ and $f_{c}geq10^3$, and does not need the spikes within the fundamental period to lie on a lattice. In this paper, we develop a theory of super-resolution from short-time Fourier transform (STFT) measurements. Specifically, we present a recovery method similar in spirit to the one in [3] for pure Fourier measurements. For a STFT Gaussian window function of width $sigma=1/(4f_c)$ this method succeeds provably if $Delta>1/f_c$, without restrictions on $f_c$. Our theory is based on a measure-theoretic formulation of the recovery problem, which leads to considerable generality in the sense of the results being grid-free and applying to spike trains on both $mathbbR$ and $mathbbT$. The case of spike trains on $mathbbR$ comes with significant technical challenges. For recovery of spike trains on $mathbbT$ we prove that the correct solution can be approximated---in weak-* topology---by solving a sequence of finite-dimensional convex programming problems.

Recommendations

• Towards a Mathematical Theory of Super‐resolution
• Efficient Fourier-Wavelet Super-Resolution
• Superresolution in the maximum entropy approach to invert Laplace transforms
• Sampling and Super Resolution of Sparse Signals Beyond the Fourier Domain
• A linear well-posed solution to recover high-frequency information for super resolution image reconstruction
• On super-resolution imaging as a multiparameter estimation problem
• Superresolution via Sparsity Constraints
• Endliche und unendliche Systeme von kontinuierlichen klassischen Orthogonalpolynomen
• Stable super-resolution of images: theoretical study

Cites work

• scientific article; zbMATH DE number 432614 (Why is no real title available?)
• scientific article; zbMATH DE number 193011 (Why is no real title available?)
• scientific article; zbMATH DE number 1022658 (Why is no real title available?)
• A theory of super-resolution from short-time Fourier transform measurements
• Analysis and harmonic synthesis
• Compressed Sensing Off the Grid
• Elements of analysis and algebra (and number theory)
• Exact reconstruction using Beurling minimal extrapolation
• Exact support recovery for sparse spikes deconvolution
• Functional analysis, Sobolev spaces and partial differential equations
• Generalized Fourier frames in terms of balayage
• Inverse problems in spaces of measures
• On the Schwartz-Bruhat space and the Paley-Wiener theorem for locally compact Abelian groups
• Positive trigonometric polynomials and signal processing applications
• Sampling Moments and Reconstructing Signals of Finite Rate of Innovation: Shannon Meets Strang–Fix
• Sampling signals with finite rate of innovation
• Signal Recovery and the Large Sieve
• Spectral theory of operators on Hilbert space
• Super-resolution by means of Beurling minimal extrapolation
• Super-resolution of point sources via convex programming
• Superresolution via Sparsity Constraints
• Techniques of variational analysis
• Towards a Mathematical Theory of Super‐resolution

Cited in

(13)

• Super-resolution by means of Beurling minimal extrapolation
• TV-based spline reconstruction with Fourier measurements: uniqueness and convergence of grid-based methods
• Towards a Mathematical Theory of Super‐resolution
• Stability and super-resolution of generalized spike recovery
• A note on spike localization for line spectrum estimation
• Sparse spikes super-resolution on thin grids II: the continuous basis pursuit
• Point source super-resolution via non-convex \(L_1\) based methods
• Super-resolution radar
• Super-Resolution Reconstruction in Frequency-Domain Optical-Coherence Tomography Using the Finite-Rate-of-Innovation Principle
• Dynamic spike superresolution and applications to ultrafast ultrasound imaging
• Spikes, roots, and aliasing: recovering bandlimited signals from roots of the short-time Fourier transform
• A theory of super-resolution from short-time Fourier transform measurements
• Beurling-type density criteria for system identification

This page was built for publication: A theory of super-resolution from short-time Fourier transform measurements

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1704865)