Abstract:Advances in compressive sensing provided reconstruction algorithms of sparse signals from linear measurements with optimal sample complexity, but natural extensions of this methodology to nonlinear inverse problems have been met with potentially fundamental sample complexity bottlenecks. In particular, tractable algorithms for compressive phase retrieval with sparsity priors have not been able to achieve optimal sample complexity. This has created an open problem in compressive phase retrieval: under generic, phaseless linear measurements, are there tractable reconstruction algorithms that succeed with optimal sample complexity? Meanwhile, progress in machine learning has led to the development of new data-driven signal priors in the form of generative models, which can outperform sparsity priors with significantly fewer measurements. In this work, we resolve the open problem in compressive phase retrieval and demonstrate that generative priors can lead to a fundamental advance by permitting optimal sample complexity by a tractable algorithm in this challenging nonlinear inverse problem. We additionally provide empirics showing that exploiting generative priors in phase retrieval can significantly outperform sparsity priors. These results provide support for generative priors as a new paradigm for signal recovery in a variety of contexts, both empirically and theoretically. The strengths of this paradigm are that (1) generative priors can represent some classes of natural signals more concisely than sparsity priors, (2) generative priors allow for direct optimization over the natural signal manifold, which is intractable under sparsity priors, and (3) the resulting non-convex optimization problems with generative priors can admit benign optimization landscapes at optimal sample complexity, perhaps surprisingly, even in cases of nonlinear measurements.

Compressed Sensing-Based Robust Phase Retrieval via Deep Generative Priors

Compressive Phase Retrieval: Optimal Sample Complexity with Deep Generative Priors

Non-propagation Fast Phase Diverse Phase Retrieval for Wavefront Measurement with Generalized FFT-based Basis Function.

Optimized Phase Retrieval Method Based on Multi-Focus Property of Zone Plate

Fast and Robust Phase Retrieval for Masked Coherent Diffractive Imaging

Low-Light Phase Retrieval With Implicit Generative Priors

Phase Retrieval with Holography and Untrained Priors: Tackling the Challenges of Low-Photon Nanoscale Imaging

Poisson-Gaussian Holographic Phase Retrieval with Score-based Image Prior

Practical Phase Retrieval Using Double Deep Image Priors

Deep Plug-and-Play HIO Approach for Phase Retrieval

Phase retrieval with semi-algebraic and ReLU neural network priors

Compressive Ptychography using Deep Image and Generative Priors

Accelerated Wirtinger Flow With Score-Based Image Priors for Holographic Phase Retrieval in Poisson-Gaussian Noise Conditions

Deep Phase Coded Image Prior

Deep Denoising Prior-Based Spectral Estimation for Phaseless Synthetic Aperture Radar

Phase retrieval and reconstruction of coherent synthesis by genetic algorithm

Unrolled Optimization with Deep Learning-based Priors for Phaseless Inverse Scattering Problems

Robust Image Retrieval From Phaseless Data

Robust phase retrieval with the swept approximate message passing (prSAMP) algorithm

PRISTA-Net: Deep Iterative Shrinkage Thresholding Network for Coded Diffraction Patterns Phase Retrieval