Harmonic Decompositions of Convolutional Networks

03/28/2020
by   Meyer Scetbon, et al.
0

We consider convolutional networks from a reproducing kernel Hilbert space viewpoint. We establish harmonic decompositions of convolutional networks, that is expansions into sums of elementary functions of increasing order. The elementary functions are related to the spherical harmonics, a fundamental class of special functions on spheres. The harmonic decompositions allow us to characterize the integral operators associated with convolutional networks, and obtain as a result statistical bounds for convolutional networks.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
11/15/2015

Computing with Harmonic Functions

This document is the manual for a free Mathematica package for computing...
research
11/23/2019

Hypergeometric functions and moments of operators

We consider some hypergeometric functions and prove that they are elemen...
research
10/21/2020

A Wigner-Eckart Theorem for Group Equivariant Convolution Kernels

Group equivariant convolutional networks (GCNNs) endow classical convolu...
research
04/20/2022

The Harmonic GBC Function Map is a Bijection if the Target Domain is Convex

Harmonic generalized barycentric coordinates (GBC) functions have been u...
research
03/25/2020

Cylindrical Convolutional Networks for Joint Object Detection and Viewpoint Estimation

Existing techniques to encode spatial invariance within deep convolution...
research
05/05/2017

Analysis and Design of Convolutional Networks via Hierarchical Tensor Decompositions

The driving force behind convolutional networks - the most successful de...
research
05/17/2016

Siamese convolutional networks based on phonetic features for cognate identification

In this paper, we explore the use of convolutional networks (ConvNets) f...

Please sign up or login with your details

Forgot password? Click here to reset