Abstract:In phase retrieval, we want to recover an unknown signal x∈Cd\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$${{\varvec{x}}}\in {{\mathbb {C}}}^d$$\end{document} from n quadratic measurements of the form yi=|⟨ai,x⟩|2+wi\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$y_i = |\langle {{\varvec{a}}}_i,{{\varvec{x}}}\rangle |^2+w_i$$\end{document}, where ai∈Cd\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$${{\varvec{a}}}_i\in {{\mathbb {C}}}^d$$\end{document} are known sensing vectors and wi\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$w_i$$\end{document} is measurement noise. We ask the following weak recovery question: What is the minimum number of measurements n needed to produce an estimator x^(y)\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$${\hat{{{\varvec{x}}}}}({{\varvec{y}}})$$\end{document} that is positively correlated with the signal x\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$${{\varvec{x}}}$$\end{document}? We consider the case of Gaussian vectors ai\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$${{\varvec{a}}}_i$$\end{document}. We prove that—in the high-dimensional limit—a sharp phase transition takes place, and we locate the threshold in the regime of vanishingly small noise. For n≤d-o(d)\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$n\le d-o(d)$$\end{document}, no estimator can do significantly better than random and achieve a strictly positive correlation. For n≥d+o(d)\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$n\ge d+o(d)$$\end{document}, a simple spectral estimator achieves a positive correlation. Surprisingly, numerical simulations with the same spectral estimator demonstrate promising performance with realistic sensing matrices. Spectral methods are used to initialize non-convex optimization algorithms in phase retrieval, and our approach can boost the performance in this setting as well. Our impossibility result is based on classical information-theoretic arguments. The spectral algorithm computes the leading eigenvector of a weighted empirical covariance matrix. We obtain a sharp characterization of the spectral properties of this random matrix using tools from free probability and generalizing a recent result by Lu and Li. Both the upper bound and lower bound generalize beyond phase retrieval to measurements yi\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$y_i$$\end{document} produced according to a generalized linear model. As a by-product of our analysis, we compare the threshold of the proposed spectral method with that of a message passing algorithm.

Support Recovery in the Phase Retrieval Model: Information-Theoretic Fundamental Limits

Signal Recovery with Partially Known Signal Support

I T ] 8 S ep 2 01 3 Sparse Recovery with Coherent Tight Frames via Analysis Dantzig Selector and Analysis LASSO ∗

Sparse Recovery with Coherent Tight Frame via Analysis Dantzig Selector and Analysis LASSO

Nonuniform Support Recovery from Noisy Random Measurements by Orthogonal Matching Pursuit.

Exact Recovery Conditions for Sparse Representations with Partial Support Information

Fundamental Limits of Weak Recovery with Applications to Phase Retrieval

Phase retrieval in high dimensions: Statistical and computational phase transitions

On unified view of nullspace-type conditions for recoveries associated with general sparsity structures

Phase Retrieval Via Nonlocal Complex-Domain Sparsity.

High-dimensional subset recovery in noise: Sparsified measurements without loss of statistical efficiency

Support Recovery of Sparse Signals from a Mixture of Linear Measurements

Sparse support recovery using correlation information in the presence of additive noise

Lower Lipschitz Bounds for Phase Retrieval from Locally Supported Measurements

Support Recovery with Orthogonal Matching Pursuit in the Presence of Noise: A New Analysis

Sparse Support Recovery with Non-smooth Loss Functions

Uncertainty Principles on Finite Abelian Groups, Restriction Theory, and Applications to Sparse Signal Recovery

Sparse Support Recovery with Phase-Only Measurements

Sparse Recovery Beyond Compressed Sensing: Separable Nonlinear Inverse Problems

Sparse Recovery for Overcomplete Frames: Sensing Matrices and Recovery Guarantees