On the Universal Approximation Property of Deep Fully Convolutional Neural Networks

11/25/2022
by   Ting Lin, et al.
0

We study the approximation of shift-invariant or equivariant functions by deep fully convolutional networks from the dynamical systems perspective. We prove that deep residual fully convolutional networks and their continuous-layer counterpart can achieve universal approximation of these symmetric functions at constant channel width. Moreover, we show that the same can be achieved by non-residual variants with at least 2 channels in each layer and convolutional kernel size of at least 2. In addition, we show that these requirements are necessary, in the sense that networks with fewer channels or smaller kernels fail to be universal approximators.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/18/2022

Deep Neural Network Approximation of Invariant Functions through Dynamical Systems

We study the approximation of functions which are invariant with respect...
research
11/18/2022

Universal Property of Convolutional Neural Networks

Universal approximation, whether a set of functions can approximate an a...
research
06/04/2023

How neural networks learn to classify chaotic time series

Neural networks are increasingly employed to model, analyze and control ...
research
02/14/2019

Fully Convolutional Networks for Text Classification

In this work I propose a new way of using fully convolutional networks f...
research
02/17/2020

Fully Convolutional Neural Networks for Raw Eye Tracking Data Segmentation, Generation, and Reconstruction

In this paper, we use fully convolutional neural networks for the semant...
research
03/20/2022

Fully Convolutional Fractional Scaling

We introduce a fully convolutional fractional scaling component, FCFS. F...
research
03/25/2020

SPFCN: Select and Prune the Fully Convolutional Networks for Real-time Parking Slot Detection

For passenger cars equipped with automatic parking function, convolution...

Please sign up or login with your details

Forgot password? Click here to reset