Learning Weakly Convex Regularizers for Convergent Image-Reconstruction Algorithms
SIAM Journal on Imaging Sciences · 31.03.2024 · 17 (1) · 91-115
DOI: 10.1137/23m1565243 ↗Kurz gefasst
Fachartikel in der Zeitschrift SIAM Journal on Imaging Sciences, online erschienen am 18.01.2024. Verfasst von Alexis Goujon und 2 weiteren Autor:innen. Die Arbeit wurde bisher 41-mal zitiert (Crossref) und nennt 57 Literaturangaben. Gefördert von: Swiss National Science Foundation, H2020 European Research Council.
Abstract (Originalsprache)
We propose to learn non-convex regularizers with a prescribed upper bound on their weak-convexity modulus. Such regularizers give rise to variational denoisers that minimize a convex energy. They rely on few parameters (less than 15,000) and offer a signal-processing interpretation as they mimic handcrafted sparsity-promoting regularizers. Through numerical experiments, we show that such denoisers outperform convex-regularization methods as well as the popular BM3D denoiser. Additionally, the learned regularizer can be deployed to solve inverse problems with iterative schemes that provably converge. For both CT and MRI reconstruction, the regularizer generalizes well and offers an excellent tradeoff between performance, number of parameters, guarantees, and interpretability when compared to other data-driven approaches.
Förderung
- Swiss National Science Foundation
- H2020 European Research Council
Zitieren
APA
Goujon, A., Neumayer, S., & Unser, M. (2024). Learning Weakly Convex Regularizers for Convergent Image-Reconstruction Algorithms. SIAM Journal on Imaging Sciences, 17(1), 91-115. https://doi.org/10.1137/23m1565243
Vancouver
Goujon A, Neumayer S, Unser M. Learning Weakly Convex Regularizers for Convergent Image-Reconstruction Algorithms. SIAM Journal on Imaging Sciences. 2024;17(1):91-115. doi:10.1137/23m1565243
BibTeX
@article{Goujon2024,
author = {Goujon, Alexis and Neumayer, Sebastian and Unser, Michael},
title = {Learning Weakly Convex Regularizers for Convergent Image-Reconstruction Algorithms},
journal = {SIAM Journal on Imaging Sciences},
year = {2024},
volume = {17},
number = {1},
pages = {91--115},
doi = {10.1137/23m1565243}
}
EUVIMED