Robust Compressive Phase Retrieval via Deep Generative Priors

TL;DR

This paper introduces a deep generative prior-based framework for phase retrieval, improving robustness and performance over traditional methods in imaging applications, including real scattering media imaging.

Contribution

It presents a novel deep generative prior approach for phase retrieval, demonstrating superior performance and robustness compared to traditional priors in various measurement settings.

Findings

Effective for Gaussian and Fourier measurements

Outperforms traditional sparsity and denoising priors

Proven on real scattering media imaging

Abstract

This paper proposes a new framework to regularize the highly ill-posed and non-linear phase retrieval problem through deep generative priors using simple gradient descent algorithm. We experimentally show effectiveness of proposed algorithm for random Gaussian measurements (practically relevant in imaging through scattering media) and Fourier friendly measurements (relevant in optical set ups). We demonstrate that proposed approach achieves impressive results when compared with traditional hand engineered priors including sparsity and denoising frameworks for number of measurements and robustness against noise. Finally, we show the effectiveness of the proposed approach on a real transmission matrix dataset in an actual application of multiple scattering media imaging.