Geodesic flow kernel for unsupervised domain adaptation
In real-world applications of visual recognition, many factors - such as pose, illumination, or image quality - can cause a significant mismatch between the source domain on which classifiers are trained and the target domain to which those classifiers are applied. As such, the classifiers often per...
Uloženo v:
| Vydáno v: | 2012 IEEE Conference on Computer Vision and Pattern Recognition s. 2066 - 2073 |
|---|---|
| Hlavní autoři: | , , , |
| Médium: | Konferenční příspěvek |
| Jazyk: | angličtina |
| Vydáno: |
IEEE
01.06.2012
|
| Témata: | |
| ISBN: | 9781467312264, 1467312266 |
| ISSN: | 1063-6919, 1063-6919 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | In real-world applications of visual recognition, many factors - such as pose, illumination, or image quality - can cause a significant mismatch between the source domain on which classifiers are trained and the target domain to which those classifiers are applied. As such, the classifiers often perform poorly on the target domain. Domain adaptation techniques aim to correct the mismatch. Existing approaches have concentrated on learning feature representations that are invariant across domains, and they often do not directly exploit low-dimensional structures that are intrinsic to many vision datasets. In this paper, we propose a new kernel-based method that takes advantage of such structures. Our geodesic flow kernel models domain shift by integrating an infinite number of subspaces that characterize changes in geometric and statistical properties from the source to the target domain. Our approach is computationally advantageous, automatically inferring important algorithmic parameters without requiring extensive cross-validation or labeled data from either domain. We also introduce a metric that reliably measures the adaptability between a pair of source and target domains. For a given target domain and several source domains, the metric can be used to automatically select the optimal source domain to adapt and avoid less desirable ones. Empirical studies on standard datasets demonstrate the advantages of our approach over competing methods. |
|---|---|
| AbstractList | In real-world applications of visual recognition, many factors - such as pose, illumination, or image quality - can cause a significant mismatch between the source domain on which classifiers are trained and the target domain to which those classifiers are applied. As such, the classifiers often perform poorly on the target domain. Domain adaptation techniques aim to correct the mismatch. Existing approaches have concentrated on learning feature representations that are invariant across domains, and they often do not directly exploit low-dimensional structures that are intrinsic to many vision datasets. In this paper, we propose a new kernel-based method that takes advantage of such structures. Our geodesic flow kernel models domain shift by integrating an infinite number of subspaces that characterize changes in geometric and statistical properties from the source to the target domain. Our approach is computationally advantageous, automatically inferring important algorithmic parameters without requiring extensive cross-validation or labeled data from either domain. We also introduce a metric that reliably measures the adaptability between a pair of source and target domains. For a given target domain and several source domains, the metric can be used to automatically select the optimal source domain to adapt and avoid less desirable ones. Empirical studies on standard datasets demonstrate the advantages of our approach over competing methods. |
| Author | Yuan Shi Fei Sha Grauman, K. Boqing Gong |
| Author_xml | – sequence: 1 surname: Boqing Gong fullname: Boqing Gong email: boqinggo@usc.edu organization: Dept. of Comput. Sci., Univ. of Southern California, Los Angeles, CA, USA – sequence: 2 surname: Yuan Shi fullname: Yuan Shi email: yuanshi@usc.edu organization: Dept. of Comput. Sci., Univ. of Southern California, Los Angeles, CA, USA – sequence: 3 surname: Fei Sha fullname: Fei Sha email: feisha@usc.edu organization: Dept. of Comput. Sci., Univ. of Southern California, Los Angeles, CA, USA – sequence: 4 givenname: K. surname: Grauman fullname: Grauman, K. email: grauman@cs.utexas.edu organization: Dept. of Comput. Sci., Univ. of Texas at Austin, Austin, TX, USA |
| BookMark | eNpNkFFLwzAUhaNOcJv9AeJL_0BrbpImzaMUN4WBIurruGtuINq1pekU_70FJ3jgcB6-w3k4CzZru5YYuwKeA3B7U709PeeCg8i1UMYCnLAFKG0kCFGKUzYHrmWmLdgzllhT_jGtZv_YBUtifOeTpga3Ys7MmjpHMdSpb7qv9IOGlprUd0N6aOOhp-EzRHKp6_YY2hQd9iOOoWsv2bnHJlJyzCV7Xd29VPfZ5nH9UN1usiAUjNkOjffGFb7kjkslSUxWWIOubVkicIlFoUoDhBzJKwFmZ3k9gRqccEIu2fXvbiCibT-EPQ7f2-MH8gckBkzz |
| ContentType | Conference Proceeding |
| DBID | 6IE 6IH CBEJK RIE RIO |
| DOI | 10.1109/CVPR.2012.6247911 |
| DatabaseName | IEEE Electronic Library (IEL) Conference Proceedings IEEE Proceedings Order Plan (POP) 1998-present by volume IEEE Xplore All Conference Proceedings IEEE Electronic Library (IEL) IEEE Proceedings Order Plans (POP) 1998-present |
| DatabaseTitleList | |
| Database_xml | – sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Applied Sciences Computer Science |
| EISBN | 1467312282 1467312274 9781467312271 9781467312288 |
| EISSN | 1063-6919 |
| EndPage | 2073 |
| ExternalDocumentID | 6247911 |
| Genre | orig-research |
| GroupedDBID | 23M 29F 29O 6IE 6IH 6IK ABDPE ACGFS ALMA_UNASSIGNED_HOLDINGS CBEJK IPLJI M43 RIE RIO RNS |
| ID | FETCH-LOGICAL-i241t-ba7ff7d5f80d0343e243e4ac16c988a103a554871ea0aef4217b90ca10c1d2d23 |
| IEDL.DBID | RIE |
| ISBN | 9781467312264 1467312266 |
| ISICitedReferencesCount | 1818 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000309166202029&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 1063-6919 |
| IngestDate | Wed Aug 27 04:27:19 EDT 2025 |
| IsPeerReviewed | false |
| IsScholarly | true |
| Language | English |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-i241t-ba7ff7d5f80d0343e243e4ac16c988a103a554871ea0aef4217b90ca10c1d2d23 |
| PageCount | 8 |
| ParticipantIDs | ieee_primary_6247911 |
| PublicationCentury | 2000 |
| PublicationDate | 2012-06 |
| PublicationDateYYYYMMDD | 2012-06-01 |
| PublicationDate_xml | – month: 06 year: 2012 text: 2012-06 |
| PublicationDecade | 2010 |
| PublicationTitle | 2012 IEEE Conference on Computer Vision and Pattern Recognition |
| PublicationTitleAbbrev | CVPR |
| PublicationYear | 2012 |
| Publisher | IEEE |
| Publisher_xml | – name: IEEE |
| SSID | ssj0000781092 ssj0023720 ssj0003211698 |
| Score | 2.4932895 |
| Snippet | In real-world applications of visual recognition, many factors - such as pose, illumination, or image quality - can cause a significant mismatch between the... |
| SourceID | ieee |
| SourceType | Publisher |
| StartPage | 2066 |
| SubjectTerms | Kernel Manifolds Measurement Principal component analysis Training Vectors Visualization |
| Title | Geodesic flow kernel for unsupervised domain adaptation |
| URI | https://ieeexplore.ieee.org/document/6247911 |
| WOSCitedRecordID | wos000309166202029&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07T8MwELZKxcBUoEW85YGRtEns-jFXFAZUVQiqbpVjX6SKklRtA3-fc5IGIbEwRIovyjm6OLk7f_cg5C71VcggGQYpLpiACxUGiusEvVZrU8aEU2UQzexZTiZqPtfTFrlvcmEAoAw-g74_LbF8l9vCb5UNRMyl9om8B1KKKler2U_xRWvCGiH0Y4aejdANohD7biwl8ilYIHSkyyQvIVmE9ofY136qx7yGP5HhYDSbvvgIsLhfz_6rDUuphcad_z3_Men9pPPRaaOoTkgLslPSqe1PWn_dWyTtWzzsaV0iHyF3gG-Spqv8i77DJoMVRUOXFtm2WPsfzRZ5uPzDLDNqnFlX0H6PvI0fXkdPQd1rIViiDt8FiZFpKt0wVaELGWcQ48GNjYTVSpkoZGbonZsITGgg5ejJJDq0eMFGLnYxOyPtLM_gnNAE4kRDJAS3gM6mSpAhGgI6UjbGO-UF6XrRLNZVOY1FLZXLv8lX5MhLv4rOuibt3aaAG3JoP3fL7ea2XAPfaDinsw |
| linkProvider | IEEE |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3dT8IwEG8ImugTKhi_3YOPDra2dO0zETEiIQYJb6RrbwkRNwJM_32vY2BMfPFhSXvNblnb7e76uw9C7hKXhQzitp_ghvG5kIEvuYrRajUmYUxYWTjRjPvRYCAnEzWskPtdLAwAFM5n0HTNAsu3mcndUVlLUB4pF8i71-acBptord2JiktbE5QYoesztG2E2mEK1NVjKbBPwXyhQlWEeYmIhaiBiG32p7LPSwAUGbY64-Gr8wGjzfL5vwqxFHKoW_vfGxyRxk9AnzfciapjUoH0hNRKDdQrv-8VkrZFHra0OokeIbOAa-kl8-zLe4dlCnMPVV0vT1f5wv1qVsjDZh96lnra6sUG3G-Qt-7DqNPzy2oL_gyl-NqPdZQkkW0nMrAB4wwoXlybUBglpQ4DptvOvAlBBxoSjrZMrAKDAya01FJ2SqpplsIZ8WKgsYJQCG4AzU0ZI0NUBVQoDcU7o3NSd1MzXWwSakzLWbn4m3xLDnqjl_60_zR4viSHbiU2vlpXpLpe5nBN9s3nerZa3hT74RtZwar6 |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=2012+IEEE+Conference+on+Computer+Vision+and+Pattern+Recognition&rft.atitle=Geodesic+flow+kernel+for+unsupervised+domain+adaptation&rft.au=Boqing+Gong&rft.au=Yuan+Shi&rft.au=Fei+Sha&rft.au=Grauman%2C+K.&rft.date=2012-06-01&rft.pub=IEEE&rft.isbn=9781467312264&rft.issn=1063-6919&rft.eissn=1063-6919&rft.spage=2066&rft.epage=2073&rft_id=info:doi/10.1109%2FCVPR.2012.6247911&rft.externalDocID=6247911 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1063-6919&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1063-6919&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1063-6919&client=summon |