Title
Deep Vanishing Point Detection: Geometric priors make dataset variations vanish
Author
Lin, Y. (TU Delft Pattern Recognition and Bioinformatics)
Wiersma, R.T. (TU Delft Computer Graphics and Visualisation)
Pintea, S. (TU Delft Pattern Recognition and Bioinformatics)
Hildebrandt, K.A. (TU Delft Computer Graphics and Visualisation)
Eisemann, E. (TU Delft Computer Graphics and Visualisation)
van Gemert, J.C. (TU Delft Pattern Recognition and Bioinformatics)
Contributor
O'Conner, L. (editor)
Date
2022
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.
Subject
Hough transform
Vanishing point detection
Deep learning
Geometric priors
To reference this document use:
http://resolver.tudelft.nl/uuid:534d6977-ff59-443a-ab9d-4363a7a92bcb
DOI
https://doi.org/10.1109/CVPR52688.2022.00601
Publisher
IEEE, Piscataway
Embargo date
2023-07-01
ISBN
978-1-6654-6947-0
Source
Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Event
2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022-06-18 → 2022-06-24, New Orleans, United States
Series
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1063-6919, 2022-June
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.
Part of collection
Institutional Repository
Document type
conference paper
Rights
© 2022 Y. Lin, R.T. Wiersma, S. Pintea, K.A. Hildebrandt, E. Eisemann, J.C. van Gemert