Gauge Equivariant Convolutional Networks and the Icosahedral CNN
Ever wanted to do a convolution on a Klein Bottle? This paper defines CNNs over manifolds such that they are independent of which coordinate frame you choose. Amazingly, this then results in an efficient practical method to achieve state-of-the-art in several tasks!
https://arxiv.org/abs/1902.04615
Abstract:
The principle of equivariance to symmetry transformations enables a theoretically grounded approach to neural network architecture design. Equivariant networks have shown excellent performance and data efficiency on vision and medical imaging problems that exhibit symmetries. Here we show how this principle can be extended beyond global symmetries to local gauge transformations. This enables the development of a very general class of convolutional neural networks on manifolds that depend only on the intrinsic geometry, and which includes many popular methods from equivariant and geometric deep learning. We implement gauge equivariant CNNs for signals defined on the surface of the icosahedron, which provides a reasonable approximation of the sphere. By choosing to work with this very regular manifold, we are able to implement the gauge equivariant convolution using a single conv2d call, making it a highly scalable and practical alternative to Spherical CNNs. Using this method, we demonstrate substantial improvements over previous methods on the task of segmenting omnidirectional images and global climate patterns.
Authors: Taco S. Cohen, Maurice Weiler, Berkay Kicanaoglu, Max Welling
Видео Gauge Equivariant Convolutional Networks and the Icosahedral CNN канала Yannic Kilcher
https://arxiv.org/abs/1902.04615
Abstract:
The principle of equivariance to symmetry transformations enables a theoretically grounded approach to neural network architecture design. Equivariant networks have shown excellent performance and data efficiency on vision and medical imaging problems that exhibit symmetries. Here we show how this principle can be extended beyond global symmetries to local gauge transformations. This enables the development of a very general class of convolutional neural networks on manifolds that depend only on the intrinsic geometry, and which includes many popular methods from equivariant and geometric deep learning. We implement gauge equivariant CNNs for signals defined on the surface of the icosahedron, which provides a reasonable approximation of the sphere. By choosing to work with this very regular manifold, we are able to implement the gauge equivariant convolution using a single conv2d call, making it a highly scalable and practical alternative to Spherical CNNs. Using this method, we demonstrate substantial improvements over previous methods on the task of segmenting omnidirectional images and global climate patterns.
Authors: Taco S. Cohen, Maurice Weiler, Berkay Kicanaoglu, Max Welling
Видео Gauge Equivariant Convolutional Networks and the Icosahedral CNN канала Yannic Kilcher
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Group Equivariant CNNs beyond Roto-Translations: B-Spline CNNs on Lie Groups, Erik BekkersOur Gauge Equivariant CNNs help machines see and understand like humans to improve AI experiencesTess Smidt: "Euclidean Neural Networks for Emulating Ab Initio Calculations and Generating Atomi..."[Talk] Making Convolutional Networks Shift-Invariant Again (8/19)Learning SO(3) Equivariant Representations with Spherical CNNsHarmonic Networks: Deep Translation and Rotation EquivarianceMax Welling: Generalizing Convolutions for Deep LearningWhat's so wrong with the Axiom of Choice ?WACV18: A Rotationally-Invariant Convolution Module by Feature Map Back-RotationCNN: Convolutional Neural Networks Explained - ComputerphileAuthor Interview - Typical Decoding for Natural Language GenerationGraphSAGE - Inductive Representation Learning on Large Graphs - Paper OverviewGeneralizing Convolutions for Deep Learning05 Imperial's Deep learning course: Equivariance and InvarianceModeling Perceptual Similarity and Shift-Invariance in Deep NetworksMachine Learning PhD Survival Guide 2021 | Advice on Topic Selection, Papers, Conferences & more!GauGAN Explained!Millimeter-Wave Directed Energy Technology DemonstrationHow to Represent a Neural Network with Matrices