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

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F22%3APU147766" target="_blank" >RIV/00216305:26230/22:PU147766 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.scitepress.org/Link.aspx?doi=10.5220/0010878200003124" target="_blank" >https://www.scitepress.org/Link.aspx?doi=10.5220/0010878200003124</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.5220/0010878200003124" target="_blank" >10.5220/0010878200003124</a>

Jazyk výsledku

angličtina

Název v původním jazyce

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

Popis výsledku v původním jazyce

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.

Název v anglickém jazyce

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

Popis výsledku anglicky

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.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

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

ISSN

—

e-ISSN

—

Počet stran výsledku

8

Strana od-do

545-552

Název nakladatele

SciTePress - Science and Technology Publications

Místo vydání

Setubal

Místo konání akce

Online

Datum konání akce

6. 2. 2022

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

000777569400058