Towards Architecture-Agnostic Untrained Network Priors for Image Reconstruction with Frequency Regularization

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Liu, Yilin, Pang, Yunkui, Li, Jiang, Chen, Yong, Yap, Pew-Thian
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913436067168256
author Liu, Yilin
Pang, Yunkui
Li, Jiang
Chen, Yong
Yap, Pew-Thian
author_facet Liu, Yilin
Pang, Yunkui
Li, Jiang
Chen, Yong
Yap, Pew-Thian
contents Untrained networks inspired by deep image priors have shown promising capabilities in recovering high-quality images from noisy or partial measurements without requiring training sets. Their success is widely attributed to implicit regularization due to the spectral bias of suitable network architectures. However, the application of such network-based priors often entails superfluous architectural decisions, risks of overfitting, and lengthy optimization processes, all of which hinder their practicality. To address these challenges, we propose efficient architecture-agnostic techniques to directly modulate the spectral bias of network priors: 1) bandwidth-constrained input, 2) bandwidth-controllable upsamplers, and 3) Lipschitz-regularized convolutional layers. We show that, with just a few lines of code, we can reduce overfitting in underperforming architectures and close performance gaps with high-performing counterparts, minimizing the need for extensive architecture tuning. This makes it possible to employ a more compact model to achieve performance similar or superior to larger models while reducing runtime. Demonstrated on inpainting-like MRI reconstruction task, our results signify for the first time that architectural biases, overfitting, and runtime issues of untrained network priors can be simultaneously addressed without architectural modifications. Our code is publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2312_09988
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Towards Architecture-Agnostic Untrained Network Priors for Image Reconstruction with Frequency Regularization
Liu, Yilin
Pang, Yunkui
Li, Jiang
Chen, Yong
Yap, Pew-Thian
Image and Video Processing
Computer Vision and Pattern Recognition
Untrained networks inspired by deep image priors have shown promising capabilities in recovering high-quality images from noisy or partial measurements without requiring training sets. Their success is widely attributed to implicit regularization due to the spectral bias of suitable network architectures. However, the application of such network-based priors often entails superfluous architectural decisions, risks of overfitting, and lengthy optimization processes, all of which hinder their practicality. To address these challenges, we propose efficient architecture-agnostic techniques to directly modulate the spectral bias of network priors: 1) bandwidth-constrained input, 2) bandwidth-controllable upsamplers, and 3) Lipschitz-regularized convolutional layers. We show that, with just a few lines of code, we can reduce overfitting in underperforming architectures and close performance gaps with high-performing counterparts, minimizing the need for extensive architecture tuning. This makes it possible to employ a more compact model to achieve performance similar or superior to larger models while reducing runtime. Demonstrated on inpainting-like MRI reconstruction task, our results signify for the first time that architectural biases, overfitting, and runtime issues of untrained network priors can be simultaneously addressed without architectural modifications. Our code is publicly available.
title Towards Architecture-Agnostic Untrained Network Priors for Image Reconstruction with Frequency Regularization
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2312.09988