Structured Random Model for Fast and Robust Phase Retrieval

TL;DR

This paper introduces structured random models for phase retrieval that combine computational efficiency with theoretical robustness, enabling fast and reliable imaging reconstructions.

Contribution

The authors propose a novel structured random model that emulates i.i.d. matrices with fewer computations, bridging the gap between Fourier-based and fully random models.

Findings

Robust reconstructions comparable to fully random models using gradient descent and spectral methods.

Structured models require at least two layers to emulate i.i.d. properties.

Method is suitable for optical implementation and practical imaging applications.

Abstract

Phase retrieval, a nonlinear problem prevalent in imaging applications, has been extensively studied using random models, some of which with i.i.d. sensing matrix components. While these models offer robust reconstruction guarantees, they are computationally expensive and impractical for real-world scenarios. In contrast, Fourier-based models, common in applications such as ptychography and coded diffraction imaging, are computationally more efficient but lack the theoretical guarantees of random models. Here, we introduce structured random models for phase retrieval that combine the efficiency of fast Fourier transforms with the versatility of random diagonal matrices. These models emulate i.i.d. random matrices at a fraction of the computational cost. Our approach demonstrates robust reconstructions comparable to fully random models using gradient descent and spectral methods. Furthermore, we establish that a minimum of two structured layers is necessary to achieve these structured-random properties. The proposed method is suitable for optical implementation and offers an efficient and robust alternative for phase retrieval in practical imaging applications.