When deep denoising meets iterative phase retrieval

Yaotian Wang, Xiaohang Sun, Jason Fleischer
Proceedings of the 37th International Conference on Machine Learning, PMLR 119:10007-10017, 2020.

Abstract

Recovering a signal from its Fourier intensity underlies many important applications, including lensless imaging and imaging through scattering media. Conventional algorithms for retrieving the phase suffer when noise is present but display global convergence when given clean data. Neural networks have been used to improve algorithm robustness, but efforts to date are sensitive to initial conditions and give inconsistent performance. Here, we combine iterative methods from phase retrieval with image statistics from deep denoisers, via regularization-by-denoising. The resulting methods inherit the advantages of each approach and outperform other noise-robust phase retrieval algorithms. Our work paves the way for hybrid imaging methods that integrate machine-learned constraints in conventional algorithms.

Cite this Paper

BibTeX
@InProceedings{pmlr-v119-wang20r, title = {When deep denoising meets iterative phase retrieval}, author = {Wang, Yaotian and Sun, Xiaohang and Fleischer, Jason}, booktitle = {Proceedings of the 37th International Conference on Machine Learning}, pages = {10007--10017}, year = {2020}, editor = {III, Hal Daumé and Singh, Aarti}, volume = {119}, series = {Proceedings of Machine Learning Research}, month = {13--18 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v119/wang20r/wang20r.pdf}, url = {https://proceedings.mlr.press/v119/wang20r.html}, abstract = {Recovering a signal from its Fourier intensity underlies many important applications, including lensless imaging and imaging through scattering media. Conventional algorithms for retrieving the phase suffer when noise is present but display global convergence when given clean data. Neural networks have been used to improve algorithm robustness, but efforts to date are sensitive to initial conditions and give inconsistent performance. Here, we combine iterative methods from phase retrieval with image statistics from deep denoisers, via regularization-by-denoising. The resulting methods inherit the advantages of each approach and outperform other noise-robust phase retrieval algorithms. Our work paves the way for hybrid imaging methods that integrate machine-learned constraints in conventional algorithms.} }
Endnote
%0 Conference Paper %T When deep denoising meets iterative phase retrieval %A Yaotian Wang %A Xiaohang Sun %A Jason Fleischer %B Proceedings of the 37th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2020 %E Hal Daumé III %E Aarti Singh %F pmlr-v119-wang20r %I PMLR %P 10007--10017 %U https://proceedings.mlr.press/v119/wang20r.html %V 119 %X Recovering a signal from its Fourier intensity underlies many important applications, including lensless imaging and imaging through scattering media. Conventional algorithms for retrieving the phase suffer when noise is present but display global convergence when given clean data. Neural networks have been used to improve algorithm robustness, but efforts to date are sensitive to initial conditions and give inconsistent performance. Here, we combine iterative methods from phase retrieval with image statistics from deep denoisers, via regularization-by-denoising. The resulting methods inherit the advantages of each approach and outperform other noise-robust phase retrieval algorithms. Our work paves the way for hybrid imaging methods that integrate machine-learned constraints in conventional algorithms.
APA
Wang, Y., Sun, X. & Fleischer, J.. (2020). When deep denoising meets iterative phase retrieval. Proceedings of the 37th International Conference on Machine Learning, in Proceedings of Machine Learning Research 119:10007-10017 Available from https://proceedings.mlr.press/v119/wang20r.html.

Related Material