Affordable Access

deepdyve-link
Publisher Website

Frames Learned by Prime Convolution Layers in a Deep Learning Framework.

Authors
  • Atto, Abdourrahmane M
  • Bisset, Rosie R
  • Trouve, Emmanuel
Type
Published Article
Journal
IEEE Transactions on Neural Networks and Learning Systems
Publisher
Institute of Electrical and Electronics Engineers
Publication Date
Jul 01, 2021
Volume
32
Issue
7
Pages
3247–3255
Identifiers
DOI: 10.1109/TNNLS.2020.3009059
PMID: 32721900
Source
Medline
Language
English
License
Unknown

Abstract

This brief addresses understandability of modern machine learning networks with respect to the statistical properties of their convolution layers. It proposes a set of tools for categorizing a convolution layer in terms of kernel property (meanlet, differencelet, or distrotlet) or kernel sequence property (frame spectra and intralayer correlation matrix). These tools are expected to be relevant for determining the generalization capabilities of a convolutional neural network. In particular, this brief highlights that the less frequency penalizing network among AlexNet, GoogleNet, RESNET101, and VGG19 is the more relevant one in terms of solutions for low-level ice-sheet feature enhancement.

Report this publication

Statistics

Seen <100 times