PuSH - Publikationsserver des Helmholtz Zentrums München

Edenhofer, G.* ; Frank, P.* ; Roth, J.* ; Leike, R.H.* ; Guerdi, M.* ; Platz, L.I. ; Guardiani, M.* ; Eberle, V.* ; Westerkamp, M.* ; Enßlin, T.A.*

Re-envisioning numerical information field theory (NIFTy.re): A library for gaussian processes and variational inference.

JOSS, DOI: 10.21105/joss.06593 (2024)
Verlagsversion DOI
Open Access Gold
Creative Commons Lizenzvertrag
maging is the process of transforming noisy, incomplete data into a space that humans can interpret. NIFTy is a Bayesian framework for imaging and has already successfully been applied to many fields in astrophysics. Previous design decisions held the performance and the development of methods in NIFTy back. We present a rewrite of NIFTy, coined NIFTy.re, which reworks the modeling principle, extends the inference strategies, and outsources much of the heavy lifting to JAX. The rewrite dramatically accelerates models written in NIFTy, lays the foundation for new types of inference machineries, improves maintainability, and enables interoperability between NIFTy and the JAX machine learning ecosystem.
Altmetric
Weitere Metriken?
[➜Einloggen]
Zusatzinfos bearbeiten [➜Einloggen]
Publikationstyp Artikel: Journalartikel
Dokumenttyp Review
Korrespondenzautor
ISSN (print) / ISBN 2475-9066
e-ISSN 2475-9066
Zeitschrift The Journal of Open Source Software
Verlag Open Journals
Nichtpatentliteratur Publikationen
Begutachtungsstatus Peer reviewed
Institut(e) Institute of Biological and Medical Imaging (IBMI)
Helmholtz Artifical Intelligence Cooperation Unit (HAICU)