Phase retrieval with physics informed zero-shot learning

TL;DR

This paper introduces a physics-informed zero-shot learning approach for phase retrieval that leverages priors learned from denoising tasks, reducing the need for large labeled datasets and improving reconstruction speed.

Contribution

It proposes a novel zero-shot learning method that uses priors from a neural network trained for denoising, combined with physics enforcement, for efficient phase retrieval.

Findings

Achieves high-quality phase retrieval with less training data.

Incorporating total variation improves reconstruction speed and quality.

Demonstrates approximately 8.5-fold faster reconstruction in experiments.

Abstract

Phase can be reliably estimated from a single diffracted intensity image, if a faithful prior information about the object is available. Examples include amplitude bounds, object support, sparsity in the spatial or a transform domain, deep image prior and the prior learnt from the labelled datasets by a deep neural network. Deep learning facilitates state of art reconstruction quality but requires a large labelled dataset (ground truth-measurement pair acquired in the same experimental conditions) for training. To alleviate this data requirement problem, this letter proposes a zero-shot learning method. The letter demonstrates that the object-prior learnt by a deep neural network while being trained for a denoising task can also be utilized for the phase retrieval, if the diffraction physics is effectively enforced on the network output. The letter additionally demonstrates that the incorporation of total variation in the proposed zero-shot framework facilitates the reconstruction of similar quality in lesser time (e.g. ~8.5 fold, for a test reported in this letter).