A Comparison of Reduced Order Models and Fourier Neural Operators for Geodesic Shooting in Diffeomorphic Registration

Authors

DOI:

https://doi.org/10.26754/jji-i3a.202613295

Abstract

Geodesic shooting in the Large Deformation Diffeomorphic Metric Mapping (LDDMM) framework is computationally expensive due to the complex dependence of the image similarity energy and the initial velocity field driving the shooting. This work compares two acceleration strategies: Proper Orthogonal Decomposition and Fourier Neural Operators for learned neural surrogates.

Downloads

Download data is not yet available.

Downloads

Published

2026-07-17

Issue

Vol. 14 (2026): Actas de la XV Jornada de Jóvenes Investigadores e Investigadoras del I3A - 9 de junio de 2026

Section

Artículos (Tecnologías de la Información y las Comunicaciones)

License

Copyright (c) 2026 Carlos Paesa Lía, Álvaro de la Asunción Deza, Mónica Hernández

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

How to Cite

Paesa Lía, C., de la Asunción Deza, Álvaro, & Hernández, M. (2026). A Comparison of Reduced Order Models and Fourier Neural Operators for Geodesic Shooting in Diffeomorphic Registration. Jornada De Jóvenes Investigadores Del I3A, 14. https://doi.org/10.26754/jji-i3a.202613295