Efficient image representation for object recognition via pivots selection
Patch-level features are essential for achieving good performance in computer vision tasks. Besides wellknown pre-defined patch-level descriptors such as scaleinvariant feature transform (SIFT) and histogram of oriented gradient (HOG), the kernel descriptor (KD) method [1] offers a new way to “grow-...
Gespeichert in:
| Veröffentlicht in: | Frontiers of Computer Science 2015-06, Vol.9 (3), p.383-391 |
|---|---|
| Hauptverfasser: | , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 391 |
|---|---|
| container_issue | 3 |
| container_start_page | 383 |
| container_title | Frontiers of Computer Science |
| container_volume | 9 |
| creator | XIE, Bojun LIU, Yi ZHANG, Hui YU, Jian |
| description | Patch-level features are essential for achieving good performance in computer vision tasks. Besides wellknown pre-defined patch-level descriptors such as scaleinvariant feature transform (SIFT) and histogram of oriented gradient (HOG), the kernel descriptor (KD) method [1] offers a new way to “grow-up” features from a match-kernel defined over image patch pairs using kernel principal component analysis (KPCA) and yields impressive results.
In this paper, we present efficient kernel descriptor (EKD) and efficient hierarchical kernel descriptor (EHKD), which are built upon incomplete Cholesky decomposition. EKD automatically selects a small number of pivot features for generating patch-level features to achieve better computational efficiency. EHKD recursively applies EKD to form image-level features layer-by-layer. Perhaps due to parsimony, we find surprisingly that the EKD and EHKD approaches achieved competitive results on several public datasets compared with other state-of-the-art methods, at an improved efficiency over KD. |
| doi_str_mv | 10.1007/s11704-015-4182-7 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_journals_2918719111</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><cqvip_id>664841049</cqvip_id><sourcerecordid>2918719111</sourcerecordid><originalsourceid>FETCH-LOGICAL-c462t-943453238356163d2ec96831f48f84c9d0963987e7a05143efd6c7396bc182a63</originalsourceid><addsrcrecordid>eNp9kMtOAjEUhhujiQR5AHcTXY_2tJ1elobgLSRudN0MnXYowSm0A4lvb3EI7lj19Jz_P5cPoVvAD4CxeEwAArMSQ1UykKQUF2hEsKpKQii_PMVEXqNJSiuMMcGkqggZofeZc9542_WF_65bW0S7iTblf9370BUuxCIsVtb0uWJC2_m_9N7XxcbvQ5-KZNe5mpM36MrV62Qnx3eMvp5nn9PXcv7x8jZ9mpeGcdKXilFWUUIlrThw2hBrFJcUHJNOMqMarDhVUlhR4woYta7hRlDFFybfVnM6RvdD300M251NvV6FXezySE0USAEKALIKBpWJIaVond7EfGH80YD1gZoeqOlMTR-oaZE9ZPCkrO1aG_87nzPJwbT07dJG2xz4Je1i6Hpv43nr3XHHZejabR55WpJzJhlgpugvXTeLfA</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2918719111</pqid></control><display><type>article</type><title>Efficient image representation for object recognition via pivots selection</title><source>Springer Nature - Complete Springer Journals</source><source>ProQuest Central</source><creator>XIE, Bojun ; LIU, Yi ; ZHANG, Hui ; YU, Jian</creator><creatorcontrib>XIE, Bojun ; LIU, Yi ; ZHANG, Hui ; YU, Jian</creatorcontrib><description>Patch-level features are essential for achieving good performance in computer vision tasks. Besides wellknown pre-defined patch-level descriptors such as scaleinvariant feature transform (SIFT) and histogram of oriented gradient (HOG), the kernel descriptor (KD) method [1] offers a new way to “grow-up” features from a match-kernel defined over image patch pairs using kernel principal component analysis (KPCA) and yields impressive results.
In this paper, we present efficient kernel descriptor (EKD) and efficient hierarchical kernel descriptor (EHKD), which are built upon incomplete Cholesky decomposition. EKD automatically selects a small number of pivot features for generating patch-level features to achieve better computational efficiency. EHKD recursively applies EKD to form image-level features layer-by-layer. Perhaps due to parsimony, we find surprisingly that the EKD and EHKD approaches achieved competitive results on several public datasets compared with other state-of-the-art methods, at an improved efficiency over KD.</description><identifier>ISSN: 2095-2228</identifier><identifier>EISSN: 2095-2236</identifier><identifier>DOI: 10.1007/s11704-015-4182-7</identifier><language>eng</language><publisher>Beijing: Higher Education Press</publisher><subject>Computer Science ; Computer vision ; efficient hierarchical kernel descriptor ; efficient kernel descriptor ; image-level features ; incomplete Cholesky decomposition ; Object recognition ; patch-level features ; Principal components analysis ; Research Article ; SIFT ; 不完全Cholesky分解 ; 图像表示 ; 对象识别 ; 描述符 ; 支点 ; 核主分量分析 ; 计算机视觉</subject><ispartof>Frontiers of Computer Science, 2015-06, Vol.9 (3), p.383-391</ispartof><rights>Copyright reserved, 2014, Higher Education Press and Springer-Verlag Berlin Heidelberg</rights><rights>Higher Education Press and Springer-Verlag Berlin Heidelberg 2015</rights><rights>Higher Education Press and Springer-Verlag Berlin Heidelberg 2015.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c462t-943453238356163d2ec96831f48f84c9d0963987e7a05143efd6c7396bc182a63</citedby><cites>FETCH-LOGICAL-c462t-943453238356163d2ec96831f48f84c9d0963987e7a05143efd6c7396bc182a63</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Uhttp://image.cqvip.com/vip1000/qk/71018X/71018X.jpg</thumbnail><linktopdf>$$Uhttps://link.springer.com/content/pdf/10.1007/s11704-015-4182-7$$EPDF$$P50$$Gspringer$$H</linktopdf><linktohtml>$$Uhttps://www.proquest.com/docview/2918719111?pq-origsite=primo$$EHTML$$P50$$Gproquest$$H</linktohtml><link.rule.ids>314,776,780,21367,27901,27902,33721,41464,42533,43781,51294</link.rule.ids></links><search><creatorcontrib>XIE, Bojun</creatorcontrib><creatorcontrib>LIU, Yi</creatorcontrib><creatorcontrib>ZHANG, Hui</creatorcontrib><creatorcontrib>YU, Jian</creatorcontrib><title>Efficient image representation for object recognition via pivots selection</title><title>Frontiers of Computer Science</title><addtitle>Front. Comput. Sci</addtitle><addtitle>Frontiers of Computer Science in China</addtitle><description>Patch-level features are essential for achieving good performance in computer vision tasks. Besides wellknown pre-defined patch-level descriptors such as scaleinvariant feature transform (SIFT) and histogram of oriented gradient (HOG), the kernel descriptor (KD) method [1] offers a new way to “grow-up” features from a match-kernel defined over image patch pairs using kernel principal component analysis (KPCA) and yields impressive results.
In this paper, we present efficient kernel descriptor (EKD) and efficient hierarchical kernel descriptor (EHKD), which are built upon incomplete Cholesky decomposition. EKD automatically selects a small number of pivot features for generating patch-level features to achieve better computational efficiency. EHKD recursively applies EKD to form image-level features layer-by-layer. Perhaps due to parsimony, we find surprisingly that the EKD and EHKD approaches achieved competitive results on several public datasets compared with other state-of-the-art methods, at an improved efficiency over KD.</description><subject>Computer Science</subject><subject>Computer vision</subject><subject>efficient hierarchical kernel descriptor</subject><subject>efficient kernel descriptor</subject><subject>image-level features</subject><subject>incomplete Cholesky decomposition</subject><subject>Object recognition</subject><subject>patch-level features</subject><subject>Principal components analysis</subject><subject>Research Article</subject><subject>SIFT</subject><subject>不完全Cholesky分解</subject><subject>图像表示</subject><subject>对象识别</subject><subject>描述符</subject><subject>支点</subject><subject>核主分量分析</subject><subject>计算机视觉</subject><issn>2095-2228</issn><issn>2095-2236</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2015</creationdate><recordtype>article</recordtype><sourceid>BENPR</sourceid><recordid>eNp9kMtOAjEUhhujiQR5AHcTXY_2tJ1elobgLSRudN0MnXYowSm0A4lvb3EI7lj19Jz_P5cPoVvAD4CxeEwAArMSQ1UykKQUF2hEsKpKQii_PMVEXqNJSiuMMcGkqggZofeZc9542_WF_65bW0S7iTblf9370BUuxCIsVtb0uWJC2_m_9N7XxcbvQ5-KZNe5mpM36MrV62Qnx3eMvp5nn9PXcv7x8jZ9mpeGcdKXilFWUUIlrThw2hBrFJcUHJNOMqMarDhVUlhR4woYta7hRlDFFybfVnM6RvdD300M251NvV6FXezySE0USAEKALIKBpWJIaVond7EfGH80YD1gZoeqOlMTR-oaZE9ZPCkrO1aG_87nzPJwbT07dJG2xz4Je1i6Hpv43nr3XHHZejabR55WpJzJhlgpugvXTeLfA</recordid><startdate>20150601</startdate><enddate>20150601</enddate><creator>XIE, Bojun</creator><creator>LIU, Yi</creator><creator>ZHANG, Hui</creator><creator>YU, Jian</creator><general>Higher Education Press</general><general>Springer Nature B.V</general><scope>2RA</scope><scope>92L</scope><scope>CQIGP</scope><scope>W92</scope><scope>~WA</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>8FE</scope><scope>8FG</scope><scope>AFKRA</scope><scope>ARAPS</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>GNUQQ</scope><scope>HCIFZ</scope><scope>JQ2</scope><scope>K7-</scope><scope>P5Z</scope><scope>P62</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope></search><sort><creationdate>20150601</creationdate><title>Efficient image representation for object recognition via pivots selection</title><author>XIE, Bojun ; LIU, Yi ; ZHANG, Hui ; YU, Jian</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c462t-943453238356163d2ec96831f48f84c9d0963987e7a05143efd6c7396bc182a63</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2015</creationdate><topic>Computer Science</topic><topic>Computer vision</topic><topic>efficient hierarchical kernel descriptor</topic><topic>efficient kernel descriptor</topic><topic>image-level features</topic><topic>incomplete Cholesky decomposition</topic><topic>Object recognition</topic><topic>patch-level features</topic><topic>Principal components analysis</topic><topic>Research Article</topic><topic>SIFT</topic><topic>不完全Cholesky分解</topic><topic>图像表示</topic><topic>对象识别</topic><topic>描述符</topic><topic>支点</topic><topic>核主分量分析</topic><topic>计算机视觉</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>XIE, Bojun</creatorcontrib><creatorcontrib>LIU, Yi</creatorcontrib><creatorcontrib>ZHANG, Hui</creatorcontrib><creatorcontrib>YU, Jian</creatorcontrib><collection>中文科技期刊数据库</collection><collection>中文科技期刊数据库-CALIS站点</collection><collection>中文科技期刊数据库-7.0平台</collection><collection>中文科技期刊数据库-工程技术</collection><collection>中文科技期刊数据库- 镜像站点</collection><collection>CrossRef</collection><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>ProQuest Central UK/Ireland</collection><collection>Advanced Technologies & Aerospace Collection</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>ProQuest Central Student</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Computer Science Collection</collection><collection>Computer Science Database</collection><collection>Advanced Technologies & Aerospace Database</collection><collection>ProQuest Advanced Technologies & Aerospace Collection</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><jtitle>Frontiers of Computer Science</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>XIE, Bojun</au><au>LIU, Yi</au><au>ZHANG, Hui</au><au>YU, Jian</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Efficient image representation for object recognition via pivots selection</atitle><jtitle>Frontiers of Computer Science</jtitle><stitle>Front. Comput. Sci</stitle><addtitle>Frontiers of Computer Science in China</addtitle><date>2015-06-01</date><risdate>2015</risdate><volume>9</volume><issue>3</issue><spage>383</spage><epage>391</epage><pages>383-391</pages><issn>2095-2228</issn><eissn>2095-2236</eissn><abstract>Patch-level features are essential for achieving good performance in computer vision tasks. Besides wellknown pre-defined patch-level descriptors such as scaleinvariant feature transform (SIFT) and histogram of oriented gradient (HOG), the kernel descriptor (KD) method [1] offers a new way to “grow-up” features from a match-kernel defined over image patch pairs using kernel principal component analysis (KPCA) and yields impressive results.
In this paper, we present efficient kernel descriptor (EKD) and efficient hierarchical kernel descriptor (EHKD), which are built upon incomplete Cholesky decomposition. EKD automatically selects a small number of pivot features for generating patch-level features to achieve better computational efficiency. EHKD recursively applies EKD to form image-level features layer-by-layer. Perhaps due to parsimony, we find surprisingly that the EKD and EHKD approaches achieved competitive results on several public datasets compared with other state-of-the-art methods, at an improved efficiency over KD.</abstract><cop>Beijing</cop><pub>Higher Education Press</pub><doi>10.1007/s11704-015-4182-7</doi><tpages>9</tpages><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 2095-2228 |
| ispartof | Frontiers of Computer Science, 2015-06, Vol.9 (3), p.383-391 |
| issn | 2095-2228 2095-2236 |
| language | eng |
| recordid | cdi_proquest_journals_2918719111 |
| source | Springer Nature - Complete Springer Journals; ProQuest Central |
| subjects | Computer Science Computer vision efficient hierarchical kernel descriptor efficient kernel descriptor image-level features incomplete Cholesky decomposition Object recognition patch-level features Principal components analysis Research Article SIFT 不完全Cholesky分解 图像表示 对象识别 描述符 支点 核主分量分析 计算机视觉 |
| title | Efficient image representation for object recognition via pivots selection |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-13T22%3A16%3A09IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Efficient%20image%20representation%20for%20object%20recognition%20via%20pivots%20selection&rft.jtitle=Frontiers%20of%20Computer%20Science&rft.au=XIE,%20Bojun&rft.date=2015-06-01&rft.volume=9&rft.issue=3&rft.spage=383&rft.epage=391&rft.pages=383-391&rft.issn=2095-2228&rft.eissn=2095-2236&rft_id=info:doi/10.1007/s11704-015-4182-7&rft_dat=%3Cproquest_cross%3E2918719111%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2918719111&rft_id=info:pmid/&rft_cqvip_id=664841049&rfr_iscdi=true |