Human eye inspired log-polar pre-processing for neural networks

Leendert A. Remmelzwaal, Amit Kumar Mishra, George F.R. Ellis

Research output: Chapter in Book/Report/Conference proceedingConference Proceeding (Non-Journal item)

Abstract

In this paper we draw inspiration from the human visual system, and present a bio-inspired pre-processing stage for neural networks. We implement this by applying a log-polar transformation as a pre-processing step, and to demonstrate, we have used a naive convolutional neural network (CNN). We demonstrate that a bio-inspired pre-processing stage can achieve rotation and scale robustness in CNNs. A key point in this paper is that the CNN does not need to be trained to identify rotation or scaling permutations; rather it is the log-polar pre-processing step that converts the image into a format that allows the CNN to handle rotation and scaling permutations. In addition we demonstrate how adding a log-polar transformation as a pre-processing step can reduce the image size to 20% of the Euclidean image size, without significantly compromising classification accuracy of the CNN. The pre-processing stage presented in this paper is modelled after the retina and therefore is only tested against an image dataset.

Original languageEnglish
Title of host publication2020 International SAUPEC/RobMech/PRASA Conference, SAUPEC/RobMech/PRASA 2020
PublisherIEEE Press
Number of pages6
ISBN (Electronic)9781728141626
DOIs
Publication statusPublished - 19 Mar 2020
Externally publishedYes
Event2020 Southern African Universities Power Engineering Conference/Robotics and Mechatronics/Pattern Recognition Association of South Africa, SAUPEC/RobMech/PRASA 2020 - Cape Town, South Africa
Duration: 29 Jan 202031 Jan 2020

Publication series

Name2020 International SAUPEC/RobMech/PRASA Conference, SAUPEC/RobMech/PRASA 2020

Conference

Conference2020 Southern African Universities Power Engineering Conference/Robotics and Mechatronics/Pattern Recognition Association of South Africa, SAUPEC/RobMech/PRASA 2020
Country/TerritorySouth Africa
CityCape Town
Period29 Jan 202031 Jan 2020

Keywords

  • Compression
  • Convolutional neural network
  • Human eye
  • Log-polar
  • Rotation invariant
  • Scale invariant

Fingerprint

Dive into the research topics of 'Human eye inspired log-polar pre-processing for neural networks'. Together they form a unique fingerprint.

Cite this