The NBNN kernel

Naive Bayes Nearest Neighbor (NBNN) has recently been proposed as a powerful, non-parametric approach for object classification, that manages to achieve remarkably good results thanks to the avoidance of a vector quantization step and the use of image-to-class comparisons, yielding good generalizati...

Full description

Saved in:
Bibliographic Details
Main Authors: Tuytelaars, T., Fritz, M., Saenko, K., Darrell, T.
Format: Conference Proceeding
Language:English
Subjects:
Accuracy
Algorithm design and analysis
Feature extraction
Kernel
Support vector machines
Training
Vectors
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 1831
container_issue
container_start_page 1824
container_title
container_volume
creator Tuytelaars, T.
Fritz, M.
Saenko, K.
Darrell, T.
description Naive Bayes Nearest Neighbor (NBNN) has recently been proposed as a powerful, non-parametric approach for object classification, that manages to achieve remarkably good results thanks to the avoidance of a vector quantization step and the use of image-to-class comparisons, yielding good generalization. In this paper, we introduce a kernelized version of NBNN. This way, we can learn the classifier in a discriminative setting. Moreover, it then becomes straightforward to combine it with other kernels. In particular, we show that our NBNN kernel is complementary to standard bag-of-features based kernels, focussing on local generalization as opposed to global image composition. By combining them, we achieve state-of-the-art results on Caltech101 and 15 Scenes datasets. As a side contribution, we also investigate how to speed up the NBNN computations.
doi_str_mv 10.1109/ICCV.2011.6126449
format conference_proceeding
fullrecord <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_6126449</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>6126449</ieee_id><sourcerecordid>6126449</sourcerecordid><originalsourceid>FETCH-LOGICAL-i175t-bf6a37765c1c4702ffd08130ba203811ec2f7fc4ace1789397a398c9864b87e83</originalsourceid><addsrcrecordid>eNo1j8tKA0EQANsXuInBs3jJD8yme149fdRFTSCsl-g1zI4zuBpFdr349wqJpzoUFBTAFWFNhLJYNc1zrZGo9qS9tXIEE7KO-c8iHUOlTUDFDu0JzITDvyN3ChU5h8pZkXOYjOMbohEdfAWXm9c8b2_bdv6eh8-8u4CzEndjnh04haf7u02zVOvHh1Vzs1Y9sftWXfHRMHuXKFlGXcoLBjLYRY0mEOWkC5dkY8rEQYxwNBKSBG-7wDmYKVzvu33Oefs19B9x-Nkexswv_Q054g</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>The NBNN kernel</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Tuytelaars, T. ; Fritz, M. ; Saenko, K. ; Darrell, T.</creator><creatorcontrib>Tuytelaars, T. ; Fritz, M. ; Saenko, K. ; Darrell, T.</creatorcontrib><description>Naive Bayes Nearest Neighbor (NBNN) has recently been proposed as a powerful, non-parametric approach for object classification, that manages to achieve remarkably good results thanks to the avoidance of a vector quantization step and the use of image-to-class comparisons, yielding good generalization. In this paper, we introduce a kernelized version of NBNN. This way, we can learn the classifier in a discriminative setting. Moreover, it then becomes straightforward to combine it with other kernels. In particular, we show that our NBNN kernel is complementary to standard bag-of-features based kernels, focussing on local generalization as opposed to global image composition. By combining them, we achieve state-of-the-art results on Caltech101 and 15 Scenes datasets. As a side contribution, we also investigate how to speed up the NBNN computations.</description><identifier>ISSN: 1550-5499</identifier><identifier>ISBN: 9781457711015</identifier><identifier>ISBN: 145771101X</identifier><identifier>EISSN: 2380-7504</identifier><identifier>EISBN: 1457711001</identifier><identifier>EISBN: 1457711028</identifier><identifier>EISBN: 9781457711022</identifier><identifier>EISBN: 9781457711008</identifier><identifier>DOI: 10.1109/ICCV.2011.6126449</identifier><language>eng</language><publisher>IEEE</publisher><subject>Accuracy ; Algorithm design and analysis ; Feature extraction ; Kernel ; Support vector machines ; Training ; Vectors</subject><ispartof>2011 International Conference on Computer Vision, 2011, p.1824-1831</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/6126449$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,2058,27925,54555,54920,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/6126449$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Tuytelaars, T.</creatorcontrib><creatorcontrib>Fritz, M.</creatorcontrib><creatorcontrib>Saenko, K.</creatorcontrib><creatorcontrib>Darrell, T.</creatorcontrib><title>The NBNN kernel</title><title>2011 International Conference on Computer Vision</title><addtitle>ICCV</addtitle><description>Naive Bayes Nearest Neighbor (NBNN) has recently been proposed as a powerful, non-parametric approach for object classification, that manages to achieve remarkably good results thanks to the avoidance of a vector quantization step and the use of image-to-class comparisons, yielding good generalization. In this paper, we introduce a kernelized version of NBNN. This way, we can learn the classifier in a discriminative setting. Moreover, it then becomes straightforward to combine it with other kernels. In particular, we show that our NBNN kernel is complementary to standard bag-of-features based kernels, focussing on local generalization as opposed to global image composition. By combining them, we achieve state-of-the-art results on Caltech101 and 15 Scenes datasets. As a side contribution, we also investigate how to speed up the NBNN computations.</description><subject>Accuracy</subject><subject>Algorithm design and analysis</subject><subject>Feature extraction</subject><subject>Kernel</subject><subject>Support vector machines</subject><subject>Training</subject><subject>Vectors</subject><issn>1550-5499</issn><issn>2380-7504</issn><isbn>9781457711015</isbn><isbn>145771101X</isbn><isbn>1457711001</isbn><isbn>1457711028</isbn><isbn>9781457711022</isbn><isbn>9781457711008</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2011</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNo1j8tKA0EQANsXuInBs3jJD8yme149fdRFTSCsl-g1zI4zuBpFdr349wqJpzoUFBTAFWFNhLJYNc1zrZGo9qS9tXIEE7KO-c8iHUOlTUDFDu0JzITDvyN3ChU5h8pZkXOYjOMbohEdfAWXm9c8b2_bdv6eh8-8u4CzEndjnh04haf7u02zVOvHh1Vzs1Y9sftWXfHRMHuXKFlGXcoLBjLYRY0mEOWkC5dkY8rEQYxwNBKSBG-7wDmYKVzvu33Oefs19B9x-Nkexswv_Q054g</recordid><startdate>201111</startdate><enddate>201111</enddate><creator>Tuytelaars, T.</creator><creator>Fritz, M.</creator><creator>Saenko, K.</creator><creator>Darrell, T.</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>201111</creationdate><title>The NBNN kernel</title><author>Tuytelaars, T. ; Fritz, M. ; Saenko, K. ; Darrell, T.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i175t-bf6a37765c1c4702ffd08130ba203811ec2f7fc4ace1789397a398c9864b87e83</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2011</creationdate><topic>Accuracy</topic><topic>Algorithm design and analysis</topic><topic>Feature extraction</topic><topic>Kernel</topic><topic>Support vector machines</topic><topic>Training</topic><topic>Vectors</topic><toplevel>online_resources</toplevel><creatorcontrib>Tuytelaars, T.</creatorcontrib><creatorcontrib>Fritz, M.</creatorcontrib><creatorcontrib>Saenko, K.</creatorcontrib><creatorcontrib>Darrell, T.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Tuytelaars, T.</au><au>Fritz, M.</au><au>Saenko, K.</au><au>Darrell, T.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>The NBNN kernel</atitle><btitle>2011 International Conference on Computer Vision</btitle><stitle>ICCV</stitle><date>2011-11</date><risdate>2011</risdate><spage>1824</spage><epage>1831</epage><pages>1824-1831</pages><issn>1550-5499</issn><eissn>2380-7504</eissn><isbn>9781457711015</isbn><isbn>145771101X</isbn><eisbn>1457711001</eisbn><eisbn>1457711028</eisbn><eisbn>9781457711022</eisbn><eisbn>9781457711008</eisbn><abstract>Naive Bayes Nearest Neighbor (NBNN) has recently been proposed as a powerful, non-parametric approach for object classification, that manages to achieve remarkably good results thanks to the avoidance of a vector quantization step and the use of image-to-class comparisons, yielding good generalization. In this paper, we introduce a kernelized version of NBNN. This way, we can learn the classifier in a discriminative setting. Moreover, it then becomes straightforward to combine it with other kernels. In particular, we show that our NBNN kernel is complementary to standard bag-of-features based kernels, focussing on local generalization as opposed to global image composition. By combining them, we achieve state-of-the-art results on Caltech101 and 15 Scenes datasets. As a side contribution, we also investigate how to speed up the NBNN computations.</abstract><pub>IEEE</pub><doi>10.1109/ICCV.2011.6126449</doi><tpages>8</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1550-5499
ispartof 2011 International Conference on Computer Vision, 2011, p.1824-1831
issn 1550-5499
2380-7504
language eng
recordid cdi_ieee_primary_6126449
source IEEE Electronic Library (IEL) Conference Proceedings
subjects Accuracy
Algorithm design and analysis
Feature extraction
Kernel
Support vector machines
Training
Vectors
title The NBNN kernel
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-21T14%3A53%3A48IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=The%20NBNN%20kernel&rft.btitle=2011%20International%20Conference%20on%20Computer%20Vision&rft.au=Tuytelaars,%20T.&rft.date=2011-11&rft.spage=1824&rft.epage=1831&rft.pages=1824-1831&rft.issn=1550-5499&rft.eissn=2380-7504&rft.isbn=9781457711015&rft.isbn_list=145771101X&rft_id=info:doi/10.1109/ICCV.2011.6126449&rft.eisbn=1457711001&rft.eisbn_list=1457711028&rft.eisbn_list=9781457711022&rft.eisbn_list=9781457711008&rft_dat=%3Cieee_6IE%3E6126449%3C/ieee_6IE%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i175t-bf6a37765c1c4702ffd08130ba203811ec2f7fc4ace1789397a398c9864b87e83%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=6126449&rfr_iscdi=true