A Unified Framework for U-Net Design and Analysis

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Williams, Christopher, Falck, Fabian, Deligiannidis, George, Holmes, Chris, Doucet, Arnaud, Syed, Saifuddin
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916086425845760
author Williams, Christopher
Falck, Fabian
Deligiannidis, George
Holmes, Chris
Doucet, Arnaud
Syed, Saifuddin
author_facet Williams, Christopher
Falck, Fabian
Deligiannidis, George
Holmes, Chris
Doucet, Arnaud
Syed, Saifuddin
contents U-Nets are a go-to, state-of-the-art neural architecture across numerous tasks for continuous signals on a square such as images and Partial Differential Equations (PDE), however their design and architecture is understudied. In this paper, we provide a framework for designing and analysing general U-Net architectures. We present theoretical results which characterise the role of the encoder and decoder in a U-Net, their high-resolution scaling limits and their conjugacy to ResNets via preconditioning. We propose Multi-ResNets, U-Nets with a simplified, wavelet-based encoder without learnable parameters. Further, we show how to design novel U-Net architectures which encode function constraints, natural bases, or the geometry of the data. In diffusion models, our framework enables us to identify that high-frequency information is dominated by noise exponentially faster, and show how U-Nets with average pooling exploit this. In our experiments, we demonstrate how Multi-ResNets achieve competitive and often superior performance compared to classical U-Nets in image segmentation, PDE surrogate modelling, and generative modelling with diffusion models. Our U-Net framework paves the way to study the theoretical properties of U-Nets and design natural, scalable neural architectures for a multitude of problems beyond the square.
format Preprint
id arxiv_https___arxiv_org_abs_2305_19638
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Unified Framework for U-Net Design and Analysis
Williams, Christopher
Falck, Fabian
Deligiannidis, George
Holmes, Chris
Doucet, Arnaud
Syed, Saifuddin
Machine Learning
Computer Vision and Pattern Recognition
Image and Video Processing
U-Nets are a go-to, state-of-the-art neural architecture across numerous tasks for continuous signals on a square such as images and Partial Differential Equations (PDE), however their design and architecture is understudied. In this paper, we provide a framework for designing and analysing general U-Net architectures. We present theoretical results which characterise the role of the encoder and decoder in a U-Net, their high-resolution scaling limits and their conjugacy to ResNets via preconditioning. We propose Multi-ResNets, U-Nets with a simplified, wavelet-based encoder without learnable parameters. Further, we show how to design novel U-Net architectures which encode function constraints, natural bases, or the geometry of the data. In diffusion models, our framework enables us to identify that high-frequency information is dominated by noise exponentially faster, and show how U-Nets with average pooling exploit this. In our experiments, we demonstrate how Multi-ResNets achieve competitive and often superior performance compared to classical U-Nets in image segmentation, PDE surrogate modelling, and generative modelling with diffusion models. Our U-Net framework paves the way to study the theoretical properties of U-Nets and design natural, scalable neural architectures for a multitude of problems beyond the square.
title A Unified Framework for U-Net Design and Analysis
topic Machine Learning
Computer Vision and Pattern Recognition
Image and Video Processing
url https://arxiv.org/abs/2305.19638