mlsanityData to 7 Oct 2026

Low Photon Count Phase Retrieval Using Deep Learning

arXiv 1806.10029 ↗First posted 25 Jun 2018eess.IV physics.opticscited by 24

Not on the map. Why?

Abstract (arXiv)

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 paper 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. Prior knowledge about the object is implicitly contained in the training data set and feature detection is possible for 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 as training it with the raw intensity measurement.

Key facts. Cited by 24 papers in the map (data through 7 Oct 2026). First posted on arXiv 25 Jun 2018.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 3; 2022: 10; 2023: 4; 2024: 5; 2025: 2.

24 papers.

Sort:

By area:

  1. 2023Physics-Informed Computer Vision: A Review and Perspectivescited by 28
  2. 2023On the use of deep learning for phase recoverycited by 20
  3. 2022Fourier Imager Network (FIN): A deep neural network for hologram reconstruction with superior external generalizationcited by 18
  4. 2022Deep learning at the edge enables real-time streaming ptychographic imagingcited by 17
  5. 2022Self-supervised learning of hologram reconstruction using physics consistencycited by 14
  6. 2021Holographic image reconstruction with phase recovery and autofocusing using recurrent neural networkscited by 10
  7. 2023eFIN: Enhanced Fourier Imager Network for generalizable autofocusing and pixel super-resolution in holographic imagingcited by 8
  8. 2022Few-shot Transfer Learning for Holographic Image Reconstruction using a Recurrent Neural Networkcited by 7
  9. 2022Power-law Scaling to Assist with Key Challenges in Artificial Intelligencecited by 6
  10. 2022Compressive Ptychography using Deep Image and Generative Priorscited by 5
  11. 2022Physics-assisted Generative Adversarial Network for X-Ray Tomographycited by 3
  12. 2024Deep-learning real-time phase retrieval of imperfect diffraction patterns from X-ray free-electron laserscited by 3
12 more
  1. 2021Deep-learning-driven Reliable Single-pixel Imaging with Uncertainty Approximationcited by 2
  2. 2022Attentional Ptycho-Tomography (APT) for three-dimensional nanoscale X-ray imaging with minimal data acquisition and computation timecited by 2
  3. 2024What is Wrong with End-to-End Learning for Phase Retrieval?cited by 2
  4. 2022Investigating the robustness of a learning-based method for quantitative phase retrieval from propagation-based x-ray phase contrast measurements under laboratory conditionscited by 1
  5. 2023Poisson-Gaussian Holographic Phase Retrieval with Score-based Image Priorcited by 1
  6. 2024Low-light phase retrieval with implicit generative priorscited by 1
  7. 2024Joint Segmentation and Image Reconstruction with Error Prediction in Photoacoustic Imaging using Deep Learningcited by 1
  8. 2021Blaschke Product Neural Networks (BPNN): A Physics-Infused Neural Network for Phase Retrieval of Meromorphic Functionscited by 0
  9. 2022A Deep Generative Approach to Oversampling in Ptychographycited by 0
  10. 20243D-2D Neural Nets for Phase Retrieval in Noisy Interferometric Imagingcited by 0
  11. 2025prNet: Data-Driven Phase Retrieval via Stochastic Refinementcited by 0
  12. 2025I2I-PR: Deep Iterative Refinement for Phase Retrieval using Image-to-Image Diffusion Modelscited by 0