Designing Rotationally Invariant Neural Networks from PDEs and Variational Methods

08/31/2021
by   Tobias Alt, et al.
17

Partial differential equation (PDE) models and their associated variational energy formulations are often rotationally invariant by design. This ensures that a rotation of the input results in a corresponding rotation of the output, which is desirable in applications such as image analysis. Convolutional neural networks (CNNs) do not share this property, and existing remedies are often complex. The goal of our paper is to investigate how diffusion and variational models achieve rotation invariance and transfer these ideas to neural networks. As a core novelty we propose activation functions which couple network channels by combining information from several oriented filters. This guarantees rotation invariance within the basic building blocks of the networks while still allowing for directional filtering. The resulting neural architectures are inherently rotationally invariant. With only a few small filters, they can achieve the same invariance as existing techniques which require a fine-grained sampling of orientations. Our findings help to translate diffusion and variational models into mathematically well-founded network architectures, and provide novel concepts for model-based CNN design.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
07/30/2021

Connections between Numerical Algorithms for PDEs and Neural Networks

We investigate numerous structural connections between numerical algorit...
research
12/07/2020

Rotation-Invariant Point Convolution With Multiple Equivariant Alignments

Recent attempts at introducing rotation invariance or equivariance in 3D...
research
03/19/2020

Local Rotation Invariance in 3D CNNs

Locally Rotation Invariant (LRI) image analysis was shown to be fundamen...
research
02/20/2020

Roto-Translation Equivariant Convolutional Networks: Application to Histopathology Image Analysis

Rotation-invariance is a desired property of machine-learning models for...
research
09/29/2022

In Search of Projectively Equivariant Neural Networks

Equivariance of linear neural network layers is well studied. In this wo...
research
04/22/2021

Equivariant Wavelets: Fast Rotation and Translation Invariant Wavelet Scattering Transforms

Wavelet scattering networks, which are convolutional neural networks (CN...
research
05/22/2021

Principled Design of Translation, Scale, and Rotation Invariant Variation Operators for Metaheuristics

In the past three decades, a large number of metaheuristics have been pr...

Please sign up or login with your details

Forgot password? Click here to reset