Publication Details

Towards Deep Learning-based 6D Bin Pose Estimation in 3D Scans

GAJDOŠECH, L.; KOCUR, V.; STUCHLÍK, M.; HUDEC, L.; MADARAS, M. Towards Deep Learning-based 6D Bin Pose Estimation in 3D Scans. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4 VISAPP: VISAPP. Setubal: SciTePress - Science and Technology Publications, 2022. p. 545-552. ISBN: 978-989-758-555-5.
Czech title
Směrem k bázi hlubokého učení 6D BIN založeného na 3D skenování
Type
conference paper
Language
English
Authors
Gajdošech Lukáš, Mgr.
Kocur Viktor, Ing., Ph.D.
Stuchlík Martin, Mgr.
Hudec Lukáš, Ing., Ph.D.
Madaras Martin, Ing., Ph.D.
URL
Keywords

Computer Vision, Bin Pose Estimation, 6D Pose Estimation, Deep Learning, Point Clouds

Abstract

An automated robotic system needs to be as robust as possible and fail-safe in general while having relatively high precision and repeatability. Although deep learning-based methods are becoming research standard on how to approach 3D scan and image processing tasks, the industry standard for processing this data is still analytically-based. Our paper claims that analytical methods are less robust and harder for testing, updating, and maintaining. This paper focuses on a specific task of 6D pose estimation of a bin in 3D scans. Therefore, we present a high-quality dataset composed of synthetic data and real scans captured by a structured-light scanner with precise annotations. Additionally, we propose two different methods for 6D bin pose estimation, an analytical method as the industrial standard and a  baseline data-driven method. Both approaches are cross-evaluated, and our experiments show that augmenting the training on real scans with synthetic data improves our proposed data-driven neural model. This position paper is preliminary, as proposed methods are trained and evaluated on a relatively small initial dataset which we plan to extend in the future.

Published
2022
Pages
545–552
Proceedings
Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4 VISAPP: VISAPP
ISBN
978-989-758-555-5
Publisher
SciTePress - Science and Technology Publications
Place
Setubal
DOI
UT WoS
000777569400058
BibTeX
@inproceedings{BUT182954,
  author="Lukáš {Gajdošech} and Viktor {Kocur} and Martin {Stuchlík} and Lukáš {Hudec} and Martin {Madaras}",
  title="Towards Deep Learning-based 6D Bin Pose Estimation in 3D Scans",
  booktitle="Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4 VISAPP: VISAPP",
  year="2022",
  pages="545--552",
  publisher="SciTePress - Science and Technology Publications",
  address="Setubal",
  doi="10.5220/0010878200003124",
  isbn="978-989-758-555-5",
  url="https://www.scitepress.org/Link.aspx?doi=10.5220/0010878200003124"
}
Back to top