Practical Phase Retrieval Using Double Deep Image Priors

TL;DR

This paper introduces a novel method for phase retrieval using double deep image priors, effectively handling the most challenging far-field PR problems without training data, outperforming existing methods significantly.

Contribution

The paper presents a new approach employing double deep image priors for far-field phase retrieval, eliminating the need for training data and reducing hyperparameter tuning.

Findings

Outperforms all competing methods by large margins

Requires no training data, enhancing practicality

Effective for the most difficult far-field PR problems

Abstract

Phase retrieval (PR) concerns the recovery of complex phases from complex magnitudes. We identify the connection between the difficulty level and the number and variety of symmetries in PR problems. We focus on the most difficult far-field PR (FFPR), and propose a novel method using double deep image priors. In realistic evaluation, our method outperforms all competing methods by large margins. As a single-instance method, our method requires no training data and minimal hyperparameter tuning, and hence enjoys good practicality.