Spectral clustering of linear subspaces for motion segmentation
Identifieur interne : 004735 ( Hal/Corpus ); précédent : 004734; suivant : 004736Spectral clustering of linear subspaces for motion segmentation
Auteurs : Fabien Lauer ; Christoph SchnörrSource :
English descriptors
Abstract
This paper studies automatic segmentation of multiple motions from tracked feature points through spectral embedding and clustering of linear subspaces. We show that the dimension of the ambient space is crucial for separability, and that low dimensions chosen in prior work are not optimal. We suggest lower and upper bounds together with a data-driven procedure for choosing the optimal ambient dimension. Application of our approach to the Hopkins155 video benchmark database uniformly outperforms a range of state-of-the-art methods both in terms of segmentation accuracy and computational speed.
Url:
Links to Exploration step
Hal:hal-00396782Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en">Spectral clustering of linear subspaces for motion segmentation</title><author><name sortKey="Lauer, Fabien" sort="Lauer, Fabien" uniqKey="Lauer F" first="Fabien" last="Lauer">Fabien Lauer</name><affiliation><hal:affiliation type="laboratory" xml:id="struct-224441" status="VALID"><orgName>Heidelberg Collaboratory for Image Processing</orgName><orgName type="acronym">HCI</orgName><desc><address><addrLine>Heidelberg Collaboratory for Image Processing, Universität Heidelberg Speyerer Straße 6 D-69115 Heidelberg</addrLine><country key="DE"></country></address><ref type="url">http://hci.iwr.uni-heidelberg.de/</ref></desc><listRelation><relation active="#struct-302846" type="direct"></relation></listRelation><tutelles><tutelle active="#struct-302846" type="direct"><org type="institution" xml:id="struct-302846" status="VALID"><orgName>Ruprecht Karls Universitat Heidelberg (GERMANY)</orgName><desc><address><country key="FR"></country></address></desc></org></tutelle></tutelles></hal:affiliation></affiliation></author><author><name sortKey="Schnorr, Christoph" sort="Schnorr, Christoph" uniqKey="Schnorr C" first="Christoph" last="Schnörr">Christoph Schnörr</name><affiliation><hal:affiliation type="laboratory" xml:id="struct-224441" status="VALID"><orgName>Heidelberg Collaboratory for Image Processing</orgName><orgName type="acronym">HCI</orgName><desc><address><addrLine>Heidelberg Collaboratory for Image Processing, Universität Heidelberg Speyerer Straße 6 D-69115 Heidelberg</addrLine><country key="DE"></country></address><ref type="url">http://hci.iwr.uni-heidelberg.de/</ref></desc><listRelation><relation active="#struct-302846" type="direct"></relation></listRelation><tutelles><tutelle active="#struct-302846" type="direct"><org type="institution" xml:id="struct-302846" status="VALID"><orgName>Ruprecht Karls Universitat Heidelberg (GERMANY)</orgName><desc><address><country key="FR"></country></address></desc></org></tutelle></tutelles></hal:affiliation></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">HAL</idno><idno type="RBID">Hal:hal-00396782</idno><idno type="halId">hal-00396782</idno><idno type="halUri">https://hal.archives-ouvertes.fr/hal-00396782</idno><idno type="url">https://hal.archives-ouvertes.fr/hal-00396782</idno><date when="2009-09">2009-09</date><idno type="wicri:Area/Hal/Corpus">004735</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en">Spectral clustering of linear subspaces for motion segmentation</title><author><name sortKey="Lauer, Fabien" sort="Lauer, Fabien" uniqKey="Lauer F" first="Fabien" last="Lauer">Fabien Lauer</name><affiliation><hal:affiliation type="laboratory" xml:id="struct-224441" status="VALID"><orgName>Heidelberg Collaboratory for Image Processing</orgName><orgName type="acronym">HCI</orgName><desc><address><addrLine>Heidelberg Collaboratory for Image Processing, Universität Heidelberg Speyerer Straße 6 D-69115 Heidelberg</addrLine><country key="DE"></country></address><ref type="url">http://hci.iwr.uni-heidelberg.de/</ref></desc><listRelation><relation active="#struct-302846" type="direct"></relation></listRelation><tutelles><tutelle active="#struct-302846" type="direct"><org type="institution" xml:id="struct-302846" status="VALID"><orgName>Ruprecht Karls Universitat Heidelberg (GERMANY)</orgName><desc><address><country key="FR"></country></address></desc></org></tutelle></tutelles></hal:affiliation></affiliation></author><author><name sortKey="Schnorr, Christoph" sort="Schnorr, Christoph" uniqKey="Schnorr C" first="Christoph" last="Schnörr">Christoph Schnörr</name><affiliation><hal:affiliation type="laboratory" xml:id="struct-224441" status="VALID"><orgName>Heidelberg Collaboratory for Image Processing</orgName><orgName type="acronym">HCI</orgName><desc><address><addrLine>Heidelberg Collaboratory for Image Processing, Universität Heidelberg Speyerer Straße 6 D-69115 Heidelberg</addrLine><country key="DE"></country></address><ref type="url">http://hci.iwr.uni-heidelberg.de/</ref></desc><listRelation><relation active="#struct-302846" type="direct"></relation></listRelation><tutelles><tutelle active="#struct-302846" type="direct"><org type="institution" xml:id="struct-302846" status="VALID"><orgName>Ruprecht Karls Universitat Heidelberg (GERMANY)</orgName><desc><address><country key="FR"></country></address></desc></org></tutelle></tutelles></hal:affiliation></affiliation></author></analytic></biblStruct></sourceDesc></fileDesc><profileDesc><textClass><keywords scheme="mix" xml:lang="en"><term>motion segmentation</term><term>spectral clustering</term><term>subspace separation</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">This paper studies automatic segmentation of multiple motions from tracked feature points through spectral embedding and clustering of linear subspaces. We show that the dimension of the ambient space is crucial for separability, and that low dimensions chosen in prior work are not optimal. We suggest lower and upper bounds together with a data-driven procedure for choosing the optimal ambient dimension. Application of our approach to the Hopkins155 video benchmark database uniformly outperforms a range of state-of-the-art methods both in terms of segmentation accuracy and computational speed.</div></front></TEI><hal api="V3"><titleStmt><title xml:lang="en">Spectral clustering of linear subspaces for motion segmentation</title><author role="aut"><persName><forename type="first">Fabien</forename><surname>Lauer</surname></persName><email>fabien.lauer@esstin.uhp-nancy.fr</email><idno type="idhal">fabien-lauer</idno><idno type="halauthor">81021</idno><affiliation ref="#struct-224441"></affiliation><affiliation ref="#struct-24433"></affiliation></author><author role="aut"><persName><forename type="first">Christoph</forename><surname>Schnörr</surname></persName><email></email><idno type="halauthor">414030</idno><affiliation ref="#struct-224441"></affiliation></author><editor role="depositor"><persName><forename>Fabien</forename><surname>Lauer</surname></persName><email>fabien.lauer@loria.fr</email></editor></titleStmt><editionStmt><edition n="v1" type="current"><date type="whenSubmitted">2009-06-18 17:54:40</date><date type="whenWritten">2009-02</date><date type="whenModified">2016-05-19 01:16:01</date><date type="whenReleased">2009-06-18 19:57:27</date><date type="whenProduced">2009-09</date><date type="whenEndEmbargoed">2009-06-18</date><ref type="file" target="https://hal.archives-ouvertes.fr/hal-00396782/document"><date notBefore="2009-06-18"></date></ref><ref type="file" subtype="author" n="1" target="https://hal.archives-ouvertes.fr/hal-00396782/file/LauerICCV09.pdf"><date notBefore="2009-06-18"></date></ref></edition><respStmt><resp>contributor</resp><name key="108844"><persName><forename>Fabien</forename><surname>Lauer</surname></persName><email>fabien.lauer@loria.fr</email></name></respStmt></editionStmt><publicationStmt><distributor>CCSD</distributor><idno type="halId">hal-00396782</idno><idno type="halUri">https://hal.archives-ouvertes.fr/hal-00396782</idno><idno type="halBibtex">lauer:hal-00396782</idno><idno type="halRefHtml">IEEE Int. Conf. on Computer Vision (ICCV'09), Sep 2009, Kyoto, Japan. 2009</idno><idno type="halRef">IEEE Int. Conf. on Computer Vision (ICCV'09), Sep 2009, Kyoto, Japan. 2009</idno></publicationStmt><seriesStmt><idno type="stamp" n="CNRS">CNRS - Centre national de la recherche scientifique</idno><idno type="stamp" n="INRIA">INRIA - Institut National de Recherche en Informatique et en Automatique</idno><idno type="stamp" n="INPL">Institut National Polytechnique de Lorraine</idno><idno type="stamp" n="LABO-LORIA-SET" p="LORIA">LABO-LORIA-SET</idno><idno type="stamp" n="LORIA">LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications</idno><idno type="stamp" n="UNIV-LORRAINE">Université de Lorraine</idno><idno type="stamp" n="LORIA2">Publications du LORIA</idno></seriesStmt><notesStmt><note type="audience" n="2">International</note><note type="invited" n="0">No</note><note type="popular" n="0">No</note><note type="peer" n="1">Yes</note><note type="proceedings" n="1">Yes</note></notesStmt><sourceDesc><biblStruct><analytic><title xml:lang="en">Spectral clustering of linear subspaces for motion segmentation</title><author role="aut"><persName><forename type="first">Fabien</forename><surname>Lauer</surname></persName><email>fabien.lauer@esstin.uhp-nancy.fr</email><idno type="idHal">fabien-lauer</idno><idno type="halAuthorId">81021</idno><affiliation ref="#struct-224441"></affiliation><affiliation ref="#struct-24433"></affiliation></author><author role="aut"><persName><forename type="first">Christoph</forename><surname>Schnörr</surname></persName><idno type="halAuthorId">414030</idno><affiliation ref="#struct-224441"></affiliation></author></analytic><monogr><title level="m">Proc. of the IEEE Int. Conf. on Computer Vision (ICCV'09)</title><meeting><title>IEEE Int. Conf. on Computer Vision (ICCV'09)</title><date type="start">2009-09</date><settlement>Kyoto</settlement><country key="JP">Japan</country></meeting><imprint><biblScope unit="pp">to appear</biblScope><date type="datePub">2009</date></imprint></monogr></biblStruct></sourceDesc><profileDesc><langUsage><language ident="en">English</language></langUsage><textClass><keywords scheme="author"><term xml:lang="en">motion segmentation</term><term xml:lang="en">spectral clustering</term><term xml:lang="en">subspace separation</term></keywords><classCode scheme="halDomain" n="info.info-cv">Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV]</classCode><classCode scheme="halDomain" n="info.info-lg">Computer Science [cs]/Machine Learning [cs.LG]</classCode><classCode scheme="halTypology" n="COMM">Conference papers</classCode></textClass><abstract xml:lang="en">This paper studies automatic segmentation of multiple motions from tracked feature points through spectral embedding and clustering of linear subspaces. We show that the dimension of the ambient space is crucial for separability, and that low dimensions chosen in prior work are not optimal. We suggest lower and upper bounds together with a data-driven procedure for choosing the optimal ambient dimension. Application of our approach to the Hopkins155 video benchmark database uniformly outperforms a range of state-of-the-art methods both in terms of segmentation accuracy and computational speed.</abstract></profileDesc></hal></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Lorraine/explor/InforLorV4/Data/Hal/Corpus
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 004735 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Hal/Corpus/biblio.hfd -nk 004735 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Lorraine
|area= InforLorV4
|flux= Hal
|étape= Corpus
|type= RBID
|clé= Hal:hal-00396782
|texte= Spectral clustering of linear subspaces for motion segmentation
}}
| This area was generated with Dilib version V0.6.33. |  |