Unsupervised Deep Learning for Phase Retrieval via Teacher-Student Distillation

Authors

  • Yuhui Quan South China University of Technology Pazhou Lab
  • Zhile Chen South China University of Technology Pazhou Lab
  • Tongyao Pang National University of Singapore
  • Hui Ji National University of Singapore

DOI:

https://doi.org/10.1609/aaai.v37i2.25306

Keywords:

CV: Low Level & Physics-Based Vision, ML: Unsupervised & Self-Supervised Learning, CV: Computational Photography, Image & Video Synthesis

Abstract

Phase retrieval (PR) is a challenging nonlinear inverse problem in scientific imaging that involves reconstructing the phase of a signal from its intensity measurements. Recently, there has been an increasing interest in deep learning-based PR. Motivated by the challenge of collecting ground-truth (GT) images in many domains, this paper proposes a fully-unsupervised learning approach for PR, which trains an end-to-end deep model via a GT-free teacher-student online distillation framework. Specifically, a teacher model is trained using a self-expressive loss with noise resistance, while a student model is trained with a consistency loss on augmented data to exploit the teacher's dark knowledge. Additionally, we develop an enhanced unfolding network for both the teacher and student models. Extensive experiments show that our proposed approach outperforms existing unsupervised PR methods with higher computational efficiency and performs competitively against supervised methods.

Downloads

Published

2023-06-26

How to Cite

Quan, Y., Chen, Z., Pang, T., & Ji, H. (2023). Unsupervised Deep Learning for Phase Retrieval via Teacher-Student Distillation. Proceedings of the AAAI Conference on Artificial Intelligence, 37(2), 2128-2136. https://doi.org/10.1609/aaai.v37i2.25306

Issue

Section

AAAI Technical Track on Computer Vision II