Low Photon Count Phase Retrieval Using Deep Learning
Name
PhysRevLett.121.243902.pdf
Size
717.65 KB
Format
Adobe PDF
Checksum (MD5)
26006bae6d72edde4c0410e17b1e6bba
Author(s)
Goy, Alexandre Sydney Robert
Arthur, Kwabena K.
Li, Shuai
Barbastathis, George
Date Issued
December 2018
Journal
Physical Review Letters
Publisher
American Physical Society
Citation
Goy, Alexandre, et al. “Low Photon Count Phase Retrieval Using Deep Learning.” Physical Review Letters, vol. 121, no. 24, Dec. 2018. © 2018 American Physical Society
Version
Final published version
Abstract
Imaging systems’ performance at low light intensity is affected by shot noise, which becomes increasingly strong as the power of the light source decreases. In this Letter, we experimentally demonstrate the use of deep neural networks to recover objects illuminated with weak light and demonstrate better performance than with the classical Gerchberg-Saxton phase retrieval algorithm for equivalent signal over noise ratio. The prior contained in the training image set can be leveraged by the deep neural network to detect features with a signal over noise ratio close to one. We apply this principle to a phase retrieval problem and show successful recovery of the object’s most salient features with as little as one photon per detector pixel on average in the illumination beam. We also show that the phase reconstruction is significantly improved by training the neural network with an initial estimate of the object, as opposed to training it with the raw intensity measurement.
MIT Department
Massachusetts Institute of Technology. Department of Mechanical Engineering
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1103/PhysRevLett.121.243902
