Generating Images with Physics-Based Rendering for an Industrial Object Detection Task: Realism versus Domain Randomization

dc.contributor.authorEversberg, Leon
dc.contributor.authorLambrecht, Jens
dc.date.accessioned2022-01-05T10:28:12Z
dc.date.available2022-01-05T10:28:12Z
dc.date.issued2021-11-26
dc.date.updated2021-12-02T16:50:31Z
dc.description.abstractLimited training data is one of the biggest challenges in the industrial application of deep learning. Generating synthetic training images is a promising solution in computer vision; however, minimizing the domain gap between synthetic and real-world images remains a problem. Therefore, based on a real-world application, we explored the generation of images with physics-based rendering for an industrial object detection task. Setting up the render engine’s environment requires a lot of choices and parameters. One fundamental question is whether to apply the concept of domain randomization or use domain knowledge to try and achieve photorealism. To answer this question, we compared different strategies for setting up lighting, background, object texture, additional foreground objects and bounding box computation in a data-centric approach. We compared the resulting average precision from generated images with different levels of realism and variability. In conclusion, we found that domain randomization is a viable strategy for the detection of industrial objects. However, domain knowledge can be used for object-related aspects to improve detection performance. Based on our results, we provide guidelines and an open-source tool for the generation of synthetic images for new industrial applications.en
dc.identifier.eissn1424-8220
dc.identifier.urihttps://depositonce.tu-berlin.de/handle/11303/16038
dc.identifier.urihttp://dx.doi.org/10.14279/depositonce-14812
dc.language.isoenen
dc.rightsLicensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/en
dc.subject.ddc620 Ingenieurwissenschaften und zugeordnete Tätigkeitende
dc.subject.otherdata-centric AIen
dc.subject.otherdeep learningen
dc.subject.otherdomain randomizationen
dc.subject.otherimage synthesisen
dc.subject.otherobject detectionen
dc.subject.otherphysics-based renderingen
dc.subject.othersynthetic imagesen
dc.titleGenerating Images with Physics-Based Rendering for an Industrial Object Detection Task: Realism versus Domain Randomizationen
dc.typeArticleen
dc.type.versionpublishedVersionen
dcterms.bibliographicCitation.articlenumber7901en
dcterms.bibliographicCitation.doi10.3390/s21237901en
dcterms.bibliographicCitation.issue23en
dcterms.bibliographicCitation.journaltitleSensorsen
dcterms.bibliographicCitation.originalpublishernameMDPIen
dcterms.bibliographicCitation.originalpublisherplaceBaselen
dcterms.bibliographicCitation.volume21en
tub.accessrights.dnbfreeen
tub.affiliationFak. 4 Elektrotechnik und Informatik>Inst. Telekommunikationssysteme>FG Industry Grade Networks and Cloudsde
tub.affiliation.facultyFak. 4 Elektrotechnik und Informatikde
tub.affiliation.groupFG Industry Grade Networks and Cloudsde
tub.affiliation.instituteInst. Telekommunikationssystemede
tub.publisher.universityorinstitutionTechnische Universität Berlinen
Files
Original bundle
Now showing 1 - 1 of 1
Loading…
Thumbnail Image
Name:
sensors-21-07901-v2.pdf
Size:
44.01 MB
Format:
Adobe Portable Document Format
Description:
Collections