Show simple item record

dc.contributor.author
Mascaro, Ruben
dc.contributor.author
Pinto Teixeira, Lucas
dc.contributor.author
Chli, Margarita
dc.date.accessioned
2021-10-27T09:32:18Z
dc.date.available
2021-05-15T18:59:48Z
dc.date.available
2021-05-17T04:56:57Z
dc.date.available
2021-05-19T09:59:39Z
dc.date.available
2021-08-03T08:27:21Z
dc.date.available
2021-10-27T09:32:18Z
dc.date.issued
2021
dc.identifier.isbn
978-1-7281-9077-8
en_US
dc.identifier.isbn
978-1-7281-9078-5
en_US
dc.identifier.other
10.1109/ICRA48506.2021.9561801
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/484229
dc.identifier.doi
10.3929/ethz-b-000484229
dc.description.abstract
Semantic 3D scene understanding is a fundamental problem in computer vision and robotics. Despite recent advances in deep learning, its application to multi-domain 3D semantic segmentation typically suffers from the lack of extensive enough annotated 3D datasets. On the contrary, 2D neural networks benefit from existing large amounts of training data and can be applied to a wider variety of environments, sometimes even without need for retraining. In this paper, we present ‘Diffuser’, a novel and efficient multi-view fusion framework that leverages 2D semantic segmentation of multiple image views of a scene to produce a consistent and refined 3D segmentation. We formulate the 3D segmentation task as a transductive label diffusion problem on a graph, where multi-view and 3D geometric properties are used to propagate semantic labels from the 2D image space to the 3D map. Experiments conducted on indoor and outdoor challenging datasets demonstrate the versatility of our approach, as well as its effectiveness for both global 3D scene labeling and single RGB-D frame segmentation. Furthermore, we show a significant increase in 3D segmentation accuracy compared to probabilistic fusion methods employed in several state-of-the-art multi-view approaches, with little computational overhead.
en_US
dc.format
application/pdf
en_US
dc.language.iso
en
en_US
dc.publisher
IEEE
en_US
dc.rights.uri
http://rightsstatements.org/page/InC-NC/1.0/
dc.subject
Semantic Scene Understanding
en_US
dc.title
Diffuser: Multi-View 2D-to-3D Label Diffusion for Semantic Scene Segmentation
en_US
dc.type
Conference Paper
dc.rights.license
In Copyright - Non-Commercial Use Permitted
dc.date.published
2021-10-18
ethz.book.title
2021 IEEE International Conference on Robotics and Automation (ICRA)
en_US
ethz.pages.start
13589
en_US
ethz.pages.end
13595
en_US
ethz.size
7 p. accepted version
en_US
ethz.version.deposit
acceptedVersion
en_US
ethz.event
2021 IEEE International Conference on Robotics and Automation (ICRA 2021)
en_US
ethz.event.location
Xi'an, China
en_US
ethz.event.date
May 30 – June 5, 2021
en_US
ethz.publication.place
Piscataway, NJ
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02130 - Dep. Maschinenbau und Verfahrenstechnik / Dep. of Mechanical and Process Eng.::02620 - Inst. f. Robotik u. Intelligente Systeme / Inst. Robotics and Intelligent Systems::09559 - Chli, Margarita (SNF-Professur) / Chli, Margarita (SNF-Professur)
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02100 - Dep. Architektur / Dep. of Architecture::02284 - NFS Digitale Fabrikation / NCCR Digital Fabrication
en_US
ethz.date.deposited
2021-05-15T19:00:15Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2021-08-03T08:27:27Z
ethz.rosetta.lastUpdated
2021-08-03T08:27:27Z
ethz.rosetta.exportRequired
true
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Diffuser:%20Multi-View%202D-to-3D%20Label%20Diffusion%20for%20Semantic%20Scene%20Segmentation&rft.date=2021&rft.spage=13589&rft.epage=13595&rft.au=Mascaro,%20Ruben&Pinto%20Teixeira,%20Lucas&Chli,%20Margarita&rft.isbn=978-1-7281-9077-8&978-1-7281-9078-5&rft.genre=proceeding&rft_id=info:doi/10.1109/ICRA48506.2021.9561801&rft.btitle=2021%20IEEE%20International%20Conference%20on%20Robotics%20and%20Automation%20(ICRA)
 Search print copy at ETH Library

Files in this item

Thumbnail

Publication type

Show simple item record