论文信息 - Covariance in Physics and Convolutional Neural Networks

Covariance in Physics and Convolutional Neural Networks

In this proceeding we give an overview of the idea of covariance (or equivariance) featured in the recent development of convolutional neural networks (CNNs). We study the similarities and differences between the use of covariance in theoretical physics and in the CNN context. Additionally, we demonstrate that the simple assumption of covariance, together with the required properties of locality, linearity and weight sharing, is sufficient to uniquely determine the form of the convolution.

[1] G. Hooft. Dimensional Reduction in Quantum Gravity , 1993, gr-qc/9310026.

[2] Max Welling,et al. Group Equivariant Convolutional Networks , 2016, ICML.

[3] Risi Kondor,et al. On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups , 2018, ICML.

[4] Taco Cohen,et al. 3D G-CNNs for Pulmonary Nodule Detection , 2018, ArXiv.

[5] Maurice Weiler,et al. Learning Steerable Filters for Rotation Equivariant CNNs , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[6] Stephan J. Garbin,et al. Harmonic Networks: Deep Translation and Rotation Equivariance , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[7] A. Einstein. The Foundation of the General Theory of Relativity , 1916 .

[8] Max Welling,et al. Spherical CNNs , 2018, ICLR.

[9] Max Welling,et al. Gauge Equivariant Convolutional Networks and the Icosahedral CNN 1 , 2019 .

[10] Gabriel J. Brostow,et al. CubeNet: Equivariance to 3D Rotation and Translation , 2018, ECCV.

[11] Max Welling,et al. 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data , 2018, NeurIPS.

[12] Maurice Weiler,et al. Intertwiners between Induced Representations (with Applications to the Theory of Equivariant Neural Networks) , 2018, ArXiv.

[13] Max Welling,et al. Steerable CNNs , 2016, ICLR.