Phase Retrieval Under a Generative Prior

07/11/2018
by   Paul Hand, et al.
0

The phase retrieval problem asks to recover a natural signal y_0 ∈R^n from m quadratic observations, where m is to be minimized. As is common in many imaging problems, natural signals are considered sparse with respect to a known basis, and the generic sparsity prior is enforced via ℓ_1 regularization. While successful in the realm of linear inverse problems, such ℓ_1 methods have encountered possibly fundamental limitations, as no computationally efficient algorithm for phase retrieval of a k-sparse signal has been proven to succeed with fewer than O(k^2 n) generic measurements, exceeding the theoretical optimum of O(k n). In this paper, we propose a novel framework for phase retrieval by 1) modeling natural signals as being in the range of a deep generative neural network G : R^k →R^n and 2) enforcing this prior directly by optimizing an empirical risk objective over the domain of the generator. Our formulation has provably favorable global geometry for gradient methods, as soon as m = O(kd^2 n), where d is the depth of the network. Specifically, when suitable deterministic conditions on the generator and measurement matrix are met, we construct a descent direction for any point outside of a small neighborhood around the unique global minimizer and its negative multiple, and show that such conditions hold with high probability under Gaussian ensembles of multilayer fully-connected generator networks and measurement matrices. This formulation for structured phase retrieval thus has two advantages over sparsity based methods: 1) deep generative priors can more tightly represent natural signals and 2) information theoretically optimal sample complexity. We corroborate these results with experiments showing that exploiting generative models in phase retrieval tasks outperforms sparse phase retrieval methods.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/24/2020

Compressive Phase Retrieval: Optimal Sample Complexity with Deep Generative Priors

Advances in compressive sensing provided reconstruction algorithms of sp...
research
07/16/2020

DeepInit Phase Retrieval

This paper shows how data-driven deep generative models can be utilized ...
research
09/09/2019

Signal retrieval with measurement system knowledge using variational generative model

Signal retrieval from a series of indirect measurements is a common task...
research
12/04/2019

Exact asymptotics for phase retrieval and compressed sensing with random generative priors

We consider the problem of compressed sensing and of (real-valued) phase...
research
05/29/2019

Global Guarantees for Blind Demodulation with Generative Priors

We study a deep learning inspired formulation for the blind demodulation...
research
02/22/2016

A Geometric Analysis of Phase Retrieval

Can we recover a complex signal from its Fourier magnitudes? More genera...
research
05/31/2022

Optimizing Intermediate Representations of Generative Models for Phase Retrieval

Phase retrieval is the problem of reconstructing images from magnitude-o...

Please sign up or login with your details

Forgot password? Click here to reset