Zur Kurzanzeige

dc.contributor.author
Kaya, Berk
dc.contributor.author
Timofte, Radu
dc.date.accessioned
2021-03-01T14:31:11Z
dc.date.available
2021-01-11T19:47:09Z
dc.date.available
2021-03-01T14:31:11Z
dc.date.issued
2020
dc.identifier.isbn
978-1-7281-8128-8
en_US
dc.identifier.isbn
978-1-7281-8129-5
en_US
dc.identifier.other
10.1109/3DV50981.2020.00114
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/461442
dc.description.abstract
We present a framework to translate between 2D image views and 3D object shapes. Recent progress in deep learning enabled us to learn structure-aware representations from a scene. However, the existing literature assumes that pairs of images and 3D shapes are available for training in full supervision. In this paper, we propose SIST, a Self-supervised Image to Shape Translation framework that fulfills three tasks: (i) reconstructing the 3D shape from a single image; (ii) learning disentangled representations for shape, appearance and viewpoint; and (iii) generating a realistic RGB image from these independent factors. In contrast to the existing approaches, our method does not require image-shape pairs for training. Instead, it uses unpaired image and shape datasets from the same object class and jointly trains image generator and shape reconstruction networks. Our translation method achieves promising results, comparable in quantitative and qualitative terms to the state-of-the-art achieved by fully-supervised methods 1 .
en_US
dc.language.iso
en
en_US
dc.publisher
IEEE
en_US
dc.subject
Shape reconstruction
en_US
dc.subject
Image generation
en_US
dc.subject
Disentagled representations
en_US
dc.subject
Self supervision
en_US
dc.subject
2D to 3D translation
en_US
dc.title
Self-Supervised 2D Image to 3D Shape Translation with Disentangled Representations
en_US
dc.type
Conference Paper
dc.date.published
2021-01-19
ethz.book.title
2020 International Conference on 3D Vision (3DV)
en_US
ethz.pages.start
1039
en_US
ethz.pages.end
1048
en_US
ethz.event
International Conference on 3D Vision (3DV 2020) (virtual)
en_US
ethz.event.location
Fukuoka, Japan
en_US
ethz.event.date
November 25-28, 2020
en_US
ethz.notes
Due to the Coronavirus (COVID-19) the conference was conducted virtually.
en_US
ethz.identifier.wos
ethz.identifier.scopus
ethz.publication.place
Piscataway, NJ
en_US
ethz.publication.status
published
en_US
ethz.date.deposited
2021-01-11T19:47:39Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Metadata only
en_US
ethz.rosetta.installDate
2021-03-01T14:31:33Z
ethz.rosetta.lastUpdated
2022-03-29T05:31:23Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Self-Supervised%202D%20Image%20to%203D%20Shape%20Translation%20with%20Disentangled%20Representations&rft.date=2020&rft.spage=1039&rft.epage=1048&rft.au=Kaya,%20Berk&Timofte,%20Radu&rft.isbn=978-1-7281-8128-8&978-1-7281-8129-5&rft.genre=proceeding&rft_id=info:doi/10.1109/3DV50981.2020.00114&rft.btitle=2020%20International%20Conference%20on%203D%20Vision%20(3DV)
 Printexemplar via ETH-Bibliothek suchen

Dateien zu diesem Eintrag

DateienGrößeFormatIm Viewer öffnen

Zu diesem Eintrag gibt es keine Dateien.

Publikationstyp

Zur Kurzanzeige