Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/35821
Full metadata record
DC FieldValueLanguage
dc.contributor.authorVANHERLE, Bram-
dc.contributor.authorPUT, Jeroen-
dc.contributor.authorMICHIELS, Nick-
dc.contributor.authorVAN REETH, Frank-
dc.date.accessioned2021-11-16T14:32:45Z-
dc.date.available2021-11-16T14:32:45Z-
dc.date.issued2021-
dc.date.submitted2021-11-09T08:46:54Z-
dc.identifier.citationGalambos, Péter; Kayacan, Erdal (Ed.). Proceedings of the 2nd International Conference on Robotics, Computer Vision and Intelligent Systems, Scitepress, p. 40 -47 (Art N° 4)-
dc.identifier.isbn9789897585371-
dc.identifier.urihttp://hdl.handle.net/1942/35821-
dc.description.abstractIn this paper a deep learning architecture is presented that can, in real time, detect the 2D locations of certain landmarks of physical tools, such as a hammer or screwdriver. To avoid the labor of manual labeling, the network is trained on synthetically generated data. Training computer vision models on computer generated images, while still achieving good accuracy on real images, is a challenge due to the difference in domain. The proposed method uses an advanced rendering method in combination with transfer learning and an intermediate supervision architecture to address this problem. It is shown that the model presented in this paper, named Intermediate Heatmap Model (IHM), generalizes to real images when trained on synthetic data. To avoid the need for an exact textured 3D model of the tool in question, it is shown that the model will generalize to an unseen tool when trained on a set of different 3D models of the same type of tool. IHM is compared to two existing approaches to keypoint detection and it is shown that it outperforms those at detecting tool landmarks, trained on synthetic data.-
dc.language.isoen-
dc.publisherScitepress-
dc.subject.otherObject Keypoint Detection-
dc.subject.otherDeep Learning-
dc.subject.otherSynthetic Data Generation-
dc.titleReal-time Detection of 2D Tool Landmarks with Synthetic Training Data-
dc.typeProceedings Paper-
local.bibliographicCitation.authorsGalambos, Péter-
local.bibliographicCitation.authorsKayacan, Erdal-
local.bibliographicCitation.conferencedate27-28 october 2021-
local.bibliographicCitation.conferencenameInternational Conference on Robotics, Computer Vision and Intelligent Systems (ROBOVIS)-
local.bibliographicCitation.conferenceplaceVirtual-
dc.identifier.epage47-
dc.identifier.spage40-
local.format.pages7-
local.bibliographicCitation.jcatC1-
local.publisher.placehttps://www.scitepress.org/PublicationsDetail.aspx?ID=rImCAGDna4o%3d&t=1-
local.type.refereedRefereed-
local.type.specifiedProceedings Paper-
local.bibliographicCitation.artnr4-
dc.identifier.doi10.5220/0010689900003061-
dc.identifier.isi000795862700004-
local.provider.typeCrossRef-
local.bibliographicCitation.btitleProceedings of the 2nd International Conference on Robotics, Computer Vision and Intelligent Systems-
local.uhasselt.uhpubyes-
local.uhasselt.internationalno-
item.fulltextWith Fulltext-
item.accessRightsOpen Access-
item.contributorVAN REETH, Frank-
item.contributorVANHERLE, Bram-
item.contributorMICHIELS, Nick-
item.contributorPUT, Jeroen-
item.fullcitationVANHERLE, Bram; PUT, Jeroen; MICHIELS, Nick & VAN REETH, Frank (2021) Real-time Detection of 2D Tool Landmarks with Synthetic Training Data. In: Galambos, Péter; Kayacan, Erdal (Ed.). Proceedings of the 2nd International Conference on Robotics, Computer Vision and Intelligent Systems, Scitepress, p. 40 -47 (Art N° 4).-
Appears in Collections:Research publications
Files in This Item:
File Description SizeFormat 
106899.pdfPublished version4.95 MBAdobe PDFView/Open
Show simple item record

Page view(s)

176
checked on May 24, 2022

Download(s)

42
checked on May 24, 2022

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.