EUVIMEDEUVIMED entdecken
Crossref

Compressive phase retrieval: Optimal sample complexity with deep generative priors

Paul Hand, Oscar Leong, Vladislav Voroninski

Communications on Pure and Applied Mathematics · 2024-02 · 77 (2) · 1147-1223

DOI: 10.1002/cpa.22155 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Communications on Pure and Applied Mathematics, online erschienen am 11.09.2023. Verfasst von Paul Hand und 2 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 81 Literaturangaben. Gefördert von: National Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Advances in compressive sensing (CS) provided reconstruction algorithms of sparse signals from linear measurements with optimal sample complexity, but natural extensions of this methodology to nonlinear inverse problems have been met with potentially fundamental sample complexity bottlenecks. In particular, tractable algorithms for compressive phase retrieval with sparsity priors have not been able to achieve optimal sample complexity. This has created an open problem in compressive phase retrieval: under generic, phaseless linear measurements, are there tractable reconstruction algorithms that succeed with optimal sample complexity? Meanwhile, progress in machine learning has led to the development of new data‐driven signal priors in the form of generative models, which can outperform sparsity priors with significantly fewer measurements. In this work, we resolve the open problem in compressive phase retrieval and demonstrate that generative priors can lead to a fundamental advance by permitting optimal sample complexity by a tractable algorithm. We additionally provide empirics showing that exploiting generative priors in phase retrieval can significantly outperform sparsity priors. These results provide support for generative priors as a new paradigm for signal recovery in a variety of contexts, both empirically and theoretically. The strengths of this paradigm are that (1) generative priors can represent some classes of natural signals more concisely than sparsity priors, (2) generative priors allow for direct optimization over the natural signal manifold, which is intractable under sparsity priors, and (3) the resulting non‐convex optimization problems with generative priors can admit benign optimization landscapes at optimal sample complexity, perhaps surprisingly, even in cases of nonlinear measurements.

Förderung

  • National Science Foundation

Zitieren

APA

Hand, P., Leong, O., & Voroninski, V. (2024). Compressive phase retrieval: Optimal sample complexity with deep generative priors. Communications on Pure and Applied Mathematics, 77(2), 1147-1223. https://doi.org/10.1002/cpa.22155

Vancouver

Hand P, Leong O, Voroninski V. Compressive phase retrieval: Optimal sample complexity with deep generative priors. Communications on Pure and Applied Mathematics. 2024;77(2):1147-1223. doi:10.1002/cpa.22155

BibTeX
@article{Hand2024,
  author = {Hand, Paul and Leong, Oscar and Voroninski, Vladislav},
  title = {Compressive phase retrieval: Optimal sample complexity with deep generative priors},
  journal = {Communications on Pure and Applied Mathematics},
  year = {2024},
  volume = {77},
  number = {2},
  pages = {1147--1223},
  doi = {10.1002/cpa.22155}
}
Kostenlos starten →