Online Architecture Search for Compressed Sensing based on Hypergradient Descent

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Nakai-Kasai, Ayano, Nakane, Yusuke, Wadayama, Tadashi
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917275728084992
author Nakai-Kasai, Ayano
Nakane, Yusuke
Wadayama, Tadashi
author_facet Nakai-Kasai, Ayano
Nakane, Yusuke
Wadayama, Tadashi
contents AS-ISTA (Architecture Searched-Iterative Shrinkage Thresholding Algorithm) and AS-FISTA (AS-Fast ISTA) are compressed sensing algorithms introducing structural parameters to ISTA and FISTA to enable architecture search within the iterative process. The structural parameters are determined using deep unfolding, but this approach requires training data and the large overhead of training time. In this paper, we propose HGD-AS-ISTA (Hypergradient Descent-AS-ISTA) and HGD-AS-FISTA that use hypergradient descent, which is an online hyperparameter optimization method, to determine the structural parameters. Experimental results show that the proposed method improves performance of the conventional ISTA/FISTA while avoiding the need for re-training when the environment changes.
format Preprint
id arxiv_https___arxiv_org_abs_2602_14411
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Online Architecture Search for Compressed Sensing based on Hypergradient Descent
Nakai-Kasai, Ayano
Nakane, Yusuke
Wadayama, Tadashi
Signal Processing
AS-ISTA (Architecture Searched-Iterative Shrinkage Thresholding Algorithm) and AS-FISTA (AS-Fast ISTA) are compressed sensing algorithms introducing structural parameters to ISTA and FISTA to enable architecture search within the iterative process. The structural parameters are determined using deep unfolding, but this approach requires training data and the large overhead of training time. In this paper, we propose HGD-AS-ISTA (Hypergradient Descent-AS-ISTA) and HGD-AS-FISTA that use hypergradient descent, which is an online hyperparameter optimization method, to determine the structural parameters. Experimental results show that the proposed method improves performance of the conventional ISTA/FISTA while avoiding the need for re-training when the environment changes.
title Online Architecture Search for Compressed Sensing based on Hypergradient Descent
topic Signal Processing
url https://arxiv.org/abs/2602.14411