Synthetic training data generation for visual object identification on load carriers

Enthalten in
Procedia CIRP 
Volume
104
Start Page
1257
End Page
1262
Citation
Procedia CIRP 104 : 1257-1262 (2021)
Contribution to Conference
54th CIRP Conference on Manufacturing Systems, CMS 2021 
Publisher DOI
10.1016/j.procir.2021.11.211
Scopus ID
2-s2.0-85121618411
Publisher
Elsevier
Peer Reviewed
true
With visual AI processes relying on individual and context accurate training data, the existing common object datasets and randomization based synthetic data pipelines can only hardly be transferred or applied on specific and narrow industrial tasks. To enable visual AI applications for intralogistics processes, such as supervision or localization of objects, a domain-knowledge driven implementation for generation of context accurate synthetic training data is introduced. With this consideration of process and domain requirements in the data generation pipeline itself, a data-generator for object identification on load carriers is contributed.
Schlagworte
Production Automation
Intralogistics
Synthetic Training Data
AI Data Generation
Object Identification
DDC Class
600: Technik
620: Ingenieurwissenschaften
Projekt(e)
Entwicklung und prototypische Erprobung von intelligenten und modularen Ladungsträgern zur schnelleren und effizienteren Materialversorgung bei der Flugzeugproduktion 
Funding Organisations
Bundesministerium für Wirtschaft und Energie - BMWi 
More Funding Information
Research was funded by the German Federal Ministry for Economics and Energy under the Program LuFo V-3 DEPOT.
Publication version
publishedVersion
Lizenz
https://creativecommons.org/licenses/by-nc-nd/4.0/
Thumbnail Image
Name

1-s2.0-S2212827121011094-main(1).pdf

Size

1.14 MB

Format

Adobe PDF