Show simple item record

dc.contributor.author
Thoma, Janine
dc.contributor.author
Paudel, Danda P.
dc.contributor.author
Van Gool, Luc
dc.contributor.editor
Larochelle, Hugo
dc.contributor.editor
Ranzato, Marc'Aurelio
dc.contributor.editor
Hadsell, Raia
dc.contributor.editor
Balcan, Maria F.
dc.contributor.editor
Lin, H.
dc.date.accessioned
2021-07-21T07:11:45Z
dc.date.available
2020-12-06T22:02:23Z
dc.date.available
2020-12-07T12:57:07Z
dc.date.available
2021-03-02T14:44:37Z
dc.date.available
2021-07-21T07:11:45Z
dc.date.issued
2021
dc.identifier.isbn
978-1-7138-2954-6
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/454834
dc.description.abstract
Localization by image retrieval is inexpensive and scalable due to simple mapping and matching techniques. Such localization, however, depends upon the quality of image features often obtained using Contrastive learning frameworks. Most contrastive learning strategies opt for features to distinguish different classes. In the context of localization, however, there is no natural definition of classes. Therefore, images are usually artificially separated into positive and negative classes, with respect to the chosen anchor images, based on some geometric proximity measure. In this paper, we show why such divisions are problematic for learning localization features. We argue that any artificial division based on some proximity measure is undesirable, due to the inherently ambiguous supervision for images near proximity threshold. To this end, we propose a novel technique that uses soft positive/negative assignments of images for contrastive learning, avoiding the aforementioned problem. Our soft assignment makes a gradual distinction between close and far images in both geometric and feature spaces. Experiments on four large-scale benchmark datasets demonstrate the superiority of the proposed soft contrastive learning over the state-of-the-art method for retrieval-based visual localization.
en_US
dc.language.iso
en
en_US
dc.publisher
Curran
en_US
dc.title
Soft Contrastive Learning for Visual Localization
en_US
dc.type
Conference Paper
dc.date.published
2020
ethz.book.title
Advances in Neural Information Processing Systems 33
en_US
ethz.pages.start
11119
en_US
ethz.pages.end
11130
en_US
ethz.event
34th Annual Conference on Neural Information Processing Systems (NeurIPS 2020)
en_US
ethz.event.location
Online
en_US
ethz.event.date
December 6-12, 2020
en_US
ethz.notes
Poster presented on December 8, 2020. Due to the Coronavirus (COVID-19) the conference was conducted virtually.
en_US
ethz.grant
ENergy aware BIM Cloud Platform in a COst-effective Building REnovation Context
en_US
ethz.publication.place
Red Hook, NY
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02652 - Institut für Bildverarbeitung / Computer Vision Laboratory::03514 - Van Gool, Luc / Van Gool, Luc
en_US
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02652 - Institut für Bildverarbeitung / Computer Vision Laboratory::03514 - Van Gool, Luc / Van Gool, Luc
en_US
ethz.identifier.url
https://papers.nips.cc/paper/2020/hash/7f2cba89a7116c7c6b0a769572d5fad9-Abstract.html
ethz.grant.agreementno
820434
ethz.grant.agreementno
820434
ethz.grant.fundername
EC
ethz.grant.fundername
EC
ethz.grant.funderDoi
10.13039/501100000780
ethz.grant.funderDoi
10.13039/501100000780
ethz.grant.program
H2020
ethz.grant.program
H2020
ethz.date.deposited
2020-12-06T22:02:38Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Metadata only
en_US
ethz.rosetta.installDate
2021-03-02T14:44:47Z
ethz.rosetta.lastUpdated
2022-03-29T10:33:19Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Soft%20Contrastive%20Learning%20for%20Visual%20Localization&rft.date=2021&rft.spage=11119&rft.epage=11130&rft.au=Thoma,%20Janine&Paudel,%20Danda%20P.&Van%20Gool,%20Luc&rft.isbn=978-1-7138-2954-6&rft.genre=proceeding&rft.btitle=Advances%20in%20Neural%20Information%20Processing%20Systems%2033
 Search print copy at ETH Library

Files in this item

FilesSizeFormatOpen in viewer

There are no files associated with this item.

Publication type

Show simple item record