Deep Vanishing Point Detection: Geometric priors make dataset variations vanish

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

Abstract

Deep learning has improved vanishing point detection in images. Yet, deep networks require expensive annotated datasets trained on costly hardware and do not generalize to even slightly different domains, and minor problem variants. Here, we address these issues by injecting deep vanishing point detection networks with prior knowledge. This prior knowledge no longer needs to be learned from data, saving valuable annotation efforts and compute, unlocking realistic few-sample scenarios, and reducing the impact of domain changes. Moreover, the interpretability of the priors allows to adapt deep networks to minor problem variations such as switching between Manhattan and non-Manhattan worlds. We seamlessly incorporate two geometric priors: (i) Hough Transform -- mapping image pixels to straight lines, and (ii) Gaussian sphere -- mapping lines to great circles whose intersections denote vanishing points. Experimentally, we ablate our choices and show comparable accuracy to existing models in the large-data setting. We validate our model's improved data efficiency, robustness to domain changes, adaptability to non-Manhattan settings.
Original languageEnglish
Title of host publicationProceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
EditorsL. O'Conner
Place of PublicationPiscataway
PublisherIEEE
Pages6093-6103
Number of pages11
ISBN (Electronic)978-1-6654-6946-3
ISBN (Print)978-1-6654-6947-0
DOIs
Publication statusPublished - 2022
Event2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) - New Orleans, United States
Duration: 18 Jun 202224 Jun 2022

Publication series

NameProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Volume2022-June
ISSN (Print)1063-6919

Conference

Conference2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Country/TerritoryUnited States
CityNew Orleans
Period18/06/2224/06/22

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

  • Hough transform
  • Vanishing point detection
  • Deep learning
  • Geometric priors

Fingerprint

Dive into the research topics of 'Deep Vanishing Point Detection: Geometric priors make dataset variations vanish'. Together they form a unique fingerprint.

Cite this