Sight-seeing in the eyes of deep neural networks

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

1 Citation (Scopus)
12 Downloads (Pure)

Abstract

We address the interpretability of convolutional neural networks (CNNs) for predicting a geo-location from an image. In a pilot experiment we classify images of Pittsburgh vs Tokyo and visualize the learned CNN filters. We found that varying the CNN architecture leads to variating in the visualized filters. This calls for further investigation of the effective parameters on the interpretability of CNNs.

Original languageEnglish
Title of host publicationProceedings - IEEE 14th International Conference on eScience, e-Science 2018
EditorsW. Hazeleger
PublisherIEEE
Pages407-408
Number of pages2
ISBN (Electronic)978-153869156-4
DOIs
Publication statusPublished - 2018
Event14th IEEE International Conference on eScience, e-Science 2018 - Amsterdam, Netherlands
Duration: 29 Oct 20181 Nov 2018

Conference

Conference14th IEEE International Conference on eScience, e-Science 2018
CountryNetherlands
CityAmsterdam
Period29/10/181/11/18

Bibliographical note

Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care

Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

  • Classification
  • Convolutional neural network (CNN)
  • Interpretability
  • Place recognition
  • Visualization

Fingerprint

Dive into the research topics of 'Sight-seeing in the eyes of deep neural networks'. Together they form a unique fingerprint.

Cite this