L-SR1: Learned Symmetric-Rank-One Preconditioning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lifshitz, Gal, Zuler, Shahar, Fouks, Ori, Raviv, Dan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911109096669184
author Lifshitz, Gal
Zuler, Shahar
Fouks, Ori
Raviv, Dan
author_facet Lifshitz, Gal
Zuler, Shahar
Fouks, Ori
Raviv, Dan
contents End-to-end deep learning has achieved impressive results but remains limited by its reliance on large labeled datasets, poor generalization to unseen scenarios, and growing computational demands. In contrast, classical optimization methods are data-efficient and lightweight but often suffer from slow convergence. While learned optimizers offer a promising fusion of both worlds, most focus on first-order methods, leaving learned second-order approaches largely unexplored. We propose a novel learned second-order optimizer that introduces a trainable preconditioning unit to enhance the classical Symmetric-Rank-One (SR1) algorithm. This unit generates data-driven vectors used to construct positive semi-definite rank-one matrices, aligned with the secant constraint via a learned projection. Our method is evaluated through analytic experiments and on the real-world task of Monocular Human Mesh Recovery (HMR), where it outperforms existing learned optimization-based approaches. Featuring a lightweight model and requiring no annotated data or fine-tuning, our approach offers strong generalization and is well-suited for integration into broader optimization-based frameworks.
format Preprint
id arxiv_https___arxiv_org_abs_2508_12270
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle L-SR1: Learned Symmetric-Rank-One Preconditioning
Lifshitz, Gal
Zuler, Shahar
Fouks, Ori
Raviv, Dan
Machine Learning
Computer Vision and Pattern Recognition
End-to-end deep learning has achieved impressive results but remains limited by its reliance on large labeled datasets, poor generalization to unseen scenarios, and growing computational demands. In contrast, classical optimization methods are data-efficient and lightweight but often suffer from slow convergence. While learned optimizers offer a promising fusion of both worlds, most focus on first-order methods, leaving learned second-order approaches largely unexplored. We propose a novel learned second-order optimizer that introduces a trainable preconditioning unit to enhance the classical Symmetric-Rank-One (SR1) algorithm. This unit generates data-driven vectors used to construct positive semi-definite rank-one matrices, aligned with the secant constraint via a learned projection. Our method is evaluated through analytic experiments and on the real-world task of Monocular Human Mesh Recovery (HMR), where it outperforms existing learned optimization-based approaches. Featuring a lightweight model and requiring no annotated data or fine-tuning, our approach offers strong generalization and is well-suited for integration into broader optimization-based frameworks.
title L-SR1: Learned Symmetric-Rank-One Preconditioning
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.12270