Analysis of non-iterative phase retrieval based on machine learning

Abstract

In this paper, we analyze a machine-learning-based non-iterative phase retrieval method. Phase retrieval and its applications have been attractive research topics in optics and photonics, for example, in biomedical imaging, astronomical imaging, and so on. Most conventional phase retrieval methods have used iterative processes to recover phase information; however, the calculation speed and convergence with these methods are serious issues in real-time monitoring applications. Machinelearning-based methods are promising for addressing these issues. Here, we numerically compare conventional methods and a machine-learning-based method in which a convolutional neural network is employed. Simulations with several conditions show that the machine-learning-based method realizes fast and robust phase recovery compared with the conventional methods. We also numerically demonstrate machine-learning-based phase retrieval from noisy measurements with a noisy training data set for improving the noise robustness. The machine-learning-based approach used in this study may increase the impact of phase retrieval, which is useful in various fields, where phase retrieval has been used as a fundamental tool.

Stepped Frequency Pulse Compression With Noncoherent Radar Using Deep Learning

“…We thereby never explicitly retrieve the phase. Since the execution time of a trained neural network is deterministic, this solution may be preferable over the aforementioned iterative solutions, an aspect that has been further studied in [12].…”

“…Although the use of deep learning has been studied in the field of phase retrieval, e.g. [12], [13], [14], and [15], it has to our knowledge not been applied in this context of non-coherent radar, and unlike most other such studies we use a DNN as a direct magnitude-to-magnitude transformation, and do not recover the phase. These two solutions are only equivalent when the phase reconstruction is perfect.…”

Non-Iterative Phase Retrieval With Cascaded Neural Networks

“…3. Non-iterative methods with a learned component: Non-iterative phase retrieval with a deep convolutional neural network that is trained end-to-end is proposed by Nishizaki et al [16]. Recently, Tayal et al [13] use symmetry breaking to solve the oversampled phase retrieval problem with neural networks.…”

“…In this section, we empirically evaluate the performance of our model. In order to do this, we report the results of the fully-convolutional residual network (ResNet) employed by Nishizaki et al [16], the multi-layer-perceptron (MLP) used in [21] and the PRCGAN [21]. In addition to these learned networks we include the results of the well-established HIO algorithm [4] and the RAAR algorithm [12] as a baseline.…”

Application of a Deep Neural Network to Phase Retrieval in Inverse Medium Scattering Problems

“…In particular, there are several works that have set out to solve phase retrieval in imaging. See, for example, [25,26]. It is worth mentioning that Xu et al proposed deep learning methods for inversion of the electromagnetic inverse scattering problem without phase information in [27].…”