Research

Paper

TESTING March 19, 2026

Factored Levenberg-Marquardt for Diffeomorphic Image Registration: An efficient optimizer for FireANTs

Authors

Rohit Jena, Pratik Chaudhari, James C. Gee

Abstract

FireANTs introduced a novel Eulerian descent method for plug-and-play behavior with arbitrary optimizers adapted for diffeomorphic image registration as a test-time optimization problem, with a GPU-accelerated implementation. FireANTs uses Adam as its default optimizer for fast and more robust optimization. However, Adam requires storing state variables (i.e. momentum and squared-momentum estimates), each of which can consume significant memory, prohibiting its use for significantly large images. In this work, we propose a modified Levenberg-Marquardt (LM) optimizer that requires only a single scalar damping parameter as optimizer state, that is adaptively tuned using a trust region approach. The resulting optimizer reduces memory by up to 24.6% for large volumes, and retaining performance across all four datasets. A single hyperparameter configuration tuned on brain MRI transfers without modification to lung CT and cross-modal abdominal registration, matching or outperforming Adam on three of four benchmarks. We also perform ablations on the effectiveness of using Metropolis-Hastings style rejection step to prevent updates that worsen the loss function.

Metadata

arXiv ID: 2603.19371
Provider: ARXIV
Primary Category: cs.CV
Published: 2026-03-19
Fetched: 2026-03-23 16:54

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.19371v1</id>\n    <title>Factored Levenberg-Marquardt for Diffeomorphic Image Registration: An efficient optimizer for FireANTs</title>\n    <updated>2026-03-19T18:04:42Z</updated>\n    <link href='https://arxiv.org/abs/2603.19371v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.19371v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>FireANTs introduced a novel Eulerian descent method for plug-and-play behavior with arbitrary optimizers adapted for diffeomorphic image registration as a test-time optimization problem, with a GPU-accelerated implementation. FireANTs uses Adam as its default optimizer for fast and more robust optimization. However, Adam requires storing state variables (i.e. momentum and squared-momentum estimates), each of which can consume significant memory, prohibiting its use for significantly large images. In this work, we propose a modified Levenberg-Marquardt (LM) optimizer that requires only a single scalar damping parameter as optimizer state, that is adaptively tuned using a trust region approach. The resulting optimizer reduces memory by up to 24.6% for large volumes, and retaining performance across all four datasets. A single hyperparameter configuration tuned on brain MRI transfers without modification to lung CT and cross-modal abdominal registration, matching or outperforming Adam on three of four benchmarks. We also perform ablations on the effectiveness of using Metropolis-Hastings style rejection step to prevent updates that worsen the loss function.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n    <published>2026-03-19T18:04:42Z</published>\n    <arxiv:primary_category term='cs.CV'/>\n    <author>\n      <name>Rohit Jena</name>\n    </author>\n    <author>\n      <name>Pratik Chaudhari</name>\n    </author>\n    <author>\n      <name>James C. Gee</name>\n    </author>\n  </entry>"
}