Notice

This record has been edited as far as possible, missing data will be added when the version of record is issued.

Show simple item record

dc.contributor.author
Ranftl, René
dc.contributor.author
Lasinger, Katrin
dc.contributor.author
Hafner, David
dc.contributor.author
Schindler, Konrad
dc.contributor.author
Koltun, Vladlen
dc.date.accessioned
2021-03-17T08:37:10Z
dc.date.available
2021-01-13T10:23:01Z
dc.date.available
2021-03-17T08:37:10Z
dc.date.issued
2020
dc.identifier.issn
0162-8828
dc.identifier.issn
1939-3539
dc.identifier.other
10.1109/TPAMI.2020.3019967
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/462024
dc.description.abstract
The success of monocular depth estimation relies on large and diverse training sets. Due to the challenges associated with acquiring dense ground-truth depth across different environments at scale, a number of datasets with distinct characteristics and biases have emerged. We develop tools that enable mixing multiple datasets during training, even if their annotations are incompatible. In particular, we propose a robust training objective that is invariant to changes in depth range and scale, advocate the use of principled multi-objective learning to combine data from different sources, and highlight the importance of pretraining encoders on auxiliary tasks. Armed with these tools, we experiment with six diverse training datasets, including a new, massive data source: 3D films. To demonstrate the generalization power of our approach we use zero-shot cross-dataset transfer, i.e. we evaluate on datasets that were not seen during training. The experiments confirm that mixing data from complementary sources greatly improves monocular depth estimation. Our approach clearly outperforms competing methods across diverse datasets, setting a new state of the art for monocular depth estimation.
en_US
dc.language.iso
en
en_US
dc.publisher
IEEE
en_US
dc.subject
Monocular depth estimation
en_US
dc.subject
Single-image depth prediction
en_US
dc.subject
Zero-shot cross-dataset transfer
en_US
dc.subject
Multi-dataset training
en_US
dc.title
Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer
en_US
dc.type
Journal Article
dc.date.published
2020-08-27
ethz.journal.title
IEEE Transactions on Pattern Analysis and Machine Intelligence
ethz.journal.abbreviated
IEEE trans. pattern anal. mach. intell.
ethz.size
14 p.
en_US
ethz.identifier.pubmed
2853149
ethz.publication.place
New York, NY
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02115 - Dep. Bau, Umwelt und Geomatik / Dep. of Civil, Env. and Geomatic Eng.::02647 - Inst. f. Geodäsie und Photogrammetrie / Institute of Geodesy and Photogrammetry::03886 - Schindler, Konrad / Schindler, Konrad
en_US
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02115 - Dep. Bau, Umwelt und Geomatik / Dep. of Civil, Env. and Geomatic Eng.::02647 - Inst. f. Geodäsie und Photogrammetrie / Institute of Geodesy and Photogrammetry::03886 - Schindler, Konrad / Schindler, Konrad
en_US
ethz.date.deposited
2021-01-13T10:23:08Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Metadata only
en_US
ethz.rosetta.exportRequired
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Towards%20Robust%20Monocular%20Depth%20Estimation:%20Mixing%20Datasets%20for%20Zero-shot%20Cross-dataset%20Transfer&rft.jtitle=IEEE%20Transactions%20on%20Pattern%20Analysis%20and%20Machine%20Intelligence&rft.date=2020&rft.issn=0162-8828&1939-3539&rft.au=Ranftl,%20Ren%C3%A9&Lasinger,%20Katrin&Hafner,%20David&Schindler,%20Konrad&Koltun,%20Vladlen&rft.genre=article&rft_id=info:doi/10.1109/TPAMI.2020.3019967&
 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