Hash Bit Selection for Nearest Neighbor Search

To overcome the barrier of storage and computation when dealing with gigantic-scale data sets, compact hashing has been studied extensively to approximate the nearest neighbor search. Despite the recent advances, critical design issues remain open in how to select the right features, hashing algorit...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	IEEE transactions on image processing Ročník 26; číslo 11; s. 5367 - 5380
Hlavní autori:	Liu, Xianglong, He, Junfeng, Chang, Shih-Fu
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	United States IEEE 01.11.2017
Predmet:	Algorithm design and analysis Binary codes bit complementarity bit reliability Dynamic programming hash bit selection Heuristic algorithms locality-sensitive hashing Nearest neighbor search Nearest neighbor searches normalized dominant set Optimization Reliability
ISSN:	1057-7149, 1941-0042, 1941-0042
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Abstract	To overcome the barrier of storage and computation when dealing with gigantic-scale data sets, compact hashing has been studied extensively to approximate the nearest neighbor search. Despite the recent advances, critical design issues remain open in how to select the right features, hashing algorithms, and/or parameter settings. In this paper, we address these by posing an optimal hash bit selection problem, in which an optimal subset of hash bits are selected from a pool of candidate bits generated by different features, algorithms, or parameters. Inspired by the optimization criteria used in existing hashing algorithms, we adopt the bit reliability and their complementarity as the selection criteria that can be carefully tailored for hashing performance in different tasks. Then, the bit selection solution is discovered by finding the best tradeoff between search accuracy and time using a modified dynamic programming method. To further reduce the computational complexity, we employ the pairwise relationship among hash bits to approximate the high-order independence property, and formulate it as an efficient quadratic programming method that is theoretically equivalent to the normalized dominant set problem in a vertex- and edge-weighted graph. Extensive large-scale experiments have been conducted under several important application scenarios of hash techniques, where our bit selection framework can achieve superior performance over both the naive selection methods and the state-of-the-art hashing algorithms, with significant accuracy gains ranging from 10% to 50%, relatively.
AbstractList	To overcome the barrier of storage and computation when dealing with gigantic-scale data sets, compact hashing has been studied extensively to approximate the nearest neighbor search. Despite the recent advances, critical design issues remain open in how to select the right features, hashing algorithms, and/or parameter settings. In this paper, we address these by posing an optimal hash bit selection problem, in which an optimal subset of hash bits are selected from a pool of candidate bits generated by different features, algorithms, or parameters. Inspired by the optimization criteria used in existing hashing algorithms, we adopt the bit reliability and their complementarity as the selection criteria that can be carefully tailored for hashing performance in different tasks. Then, the bit selection solution is discovered by finding the best tradeoff between search accuracy and time using a modified dynamic programming method. To further reduce the computational complexity, we employ the pairwise relationship among hash bits to approximate the high-order independence property, and formulate it as an efficient quadratic programming method that is theoretically equivalent to the normalized dominant set problem in a vertex- and edge-weighted graph. Extensive large-scale experiments have been conducted under several important application scenarios of hash techniques, where our bit selection framework can achieve superior performance over both the naive selection methods and the state-of-the-art hashing algorithms, with significant accuracy gains ranging from 10% to 50%, relatively. To overcome the barrier of storage and computation when dealing with gigantic-scale data sets, compact hashing has been studied extensively to approximate the nearest neighbor search. Despite the recent advances, critical design issues remain open in how to select the right features, hashing algorithms, and/or parameter settings. In this paper, we address these by posing an optimal hash bit selection problem, in which an optimal subset of hash bits are selected from a pool of candidate bits generated by different features, algorithms, or parameters. Inspired by the optimization criteria used in existing hashing algorithms, we adopt the bit reliability and their complementarity as the selection criteria that can be carefully tailored for hashing performance in different tasks. Then, the bit selection solution is discovered by finding the best tradeoff between search accuracy and time using a modified dynamic programming method. To further reduce the computational complexity, we employ the pairwise relationship among hash bits to approximate the high-order independence property, and formulate it as an efficient quadratic programming method that is theoretically equivalent to the normalized dominant set problem in a vertex- and edge-weighted graph. Extensive large-scale experiments have been conducted under several important application scenarios of hash techniques, where our bit selection framework can achieve superior performance over both the naive selection methods and the state-of-the-art hashing algorithms, with significant accuracy gains ranging from 10% to 50%, relatively.To overcome the barrier of storage and computation when dealing with gigantic-scale data sets, compact hashing has been studied extensively to approximate the nearest neighbor search. Despite the recent advances, critical design issues remain open in how to select the right features, hashing algorithms, and/or parameter settings. In this paper, we address these by posing an optimal hash bit selection problem, in which an optimal subset of hash bits are selected from a pool of candidate bits generated by different features, algorithms, or parameters. Inspired by the optimization criteria used in existing hashing algorithms, we adopt the bit reliability and their complementarity as the selection criteria that can be carefully tailored for hashing performance in different tasks. Then, the bit selection solution is discovered by finding the best tradeoff between search accuracy and time using a modified dynamic programming method. To further reduce the computational complexity, we employ the pairwise relationship among hash bits to approximate the high-order independence property, and formulate it as an efficient quadratic programming method that is theoretically equivalent to the normalized dominant set problem in a vertex- and edge-weighted graph. Extensive large-scale experiments have been conducted under several important application scenarios of hash techniques, where our bit selection framework can achieve superior performance over both the naive selection methods and the state-of-the-art hashing algorithms, with significant accuracy gains ranging from 10% to 50%, relatively.
Author	Xianglong Liu Shih-Fu Chang Junfeng He
Author_xml	– sequence: 1 givenname: Xianglong orcidid: 0000-0001-8425-4195 surname: Liu fullname: Liu, Xianglong – sequence: 2 givenname: Junfeng surname: He fullname: He, Junfeng – sequence: 3 givenname: Shih-Fu surname: Chang fullname: Chang, Shih-Fu
BackLink	https://www.ncbi.nlm.nih.gov/pubmed/28436872$$D View this record in MEDLINE/PubMed
BookMark	eNp9kL1PwzAQxS1URD9gR0JCGVkSfI6d2CNUQCtVgNTuluOcaVCalDgd-O9xaWFgYLrT0--d7r0xGTRtg4RcAk0AqLpdzV8TRiFPWKaEVOKEjEBxiCnlbBB2KvI4B66GZOz9O6XABWRnZMgkTzOZsxFJZsavo_uqj5ZYo-2rtolc20XPaDr0fZjV27oIwjIIdn1OTp2pPV4c54SsHh9W01m8eHmaT-8WsU1B9bGFEoR1HIXkCi11wmEqDS-E4gY4mJI5yei3Qg2ooqTOKpRgbJkBSyfk5nB227Ufu_CH3lTeYl2bBtud1yBViCJ5ngX0-ojuig2WettVG9N96p-IAcgOgO1a7zt02la92QftO1PVGqjed6lDl3rfpT52GYz0j_Hn9j-Wq4OlQsRfPFeUZxlLvwBAmnzS
CODEN	IIPRE4
CitedBy_id	crossref_primary_10_1186_s13640_019_0428_5 crossref_primary_10_1049_iet_ipr_2017_0770 crossref_primary_10_1007_s10489_022_03546_9 crossref_primary_10_1007_s11042_023_17810_8 crossref_primary_10_1016_j_knosys_2022_108757 crossref_primary_10_1186_s13640_019_0442_7 crossref_primary_10_1109_TIP_2019_2903661 crossref_primary_10_1109_TCYB_2018_2816791 crossref_primary_10_1109_TIP_2017_2781422 crossref_primary_10_1109_TMM_2020_3004962 crossref_primary_10_1007_s11263_020_01331_0 crossref_primary_10_1109_TNNLS_2020_3018790 crossref_primary_10_1109_TNNLS_2021_3068500 crossref_primary_10_1109_TCYB_2021_3102941 crossref_primary_10_1016_j_image_2019_115650 crossref_primary_10_1016_j_patrec_2017_11_018 crossref_primary_10_1109_TKDE_2023_3297195 crossref_primary_10_1109_TNNLS_2020_2965992 crossref_primary_10_1016_j_ins_2021_03_006 crossref_primary_10_1109_TIP_2019_2891967 crossref_primary_10_1109_TETCI_2024_3405370 crossref_primary_10_1016_j_patcog_2020_107409 crossref_primary_10_1007_s11280_018_0540_y crossref_primary_10_1109_TBDATA_2022_3161905 crossref_primary_10_1016_j_neucom_2018_06_096 crossref_primary_10_1109_TCYB_2020_2964993 crossref_primary_10_1016_j_patcog_2020_107370 crossref_primary_10_1016_j_patrec_2019_01_013 crossref_primary_10_1016_j_patcog_2020_107338 crossref_primary_10_1109_TIP_2017_2729896 crossref_primary_10_1109_TMM_2020_2991513 crossref_primary_10_1109_TIP_2018_2818008 crossref_primary_10_1016_j_patrec_2019_04_017
Cites_doi	10.1109/TCYB.2013.2289351 10.1109/TIP.2015.2505180 10.1109/CVPR.2006.130 10.1145/2540990 10.1109/TPAMI.2007.250608 10.1145/361002.361007 10.1109/CVPR.2015.7298598 10.1016/j.neucom.2014.09.033 10.1109/CVPR.2015.7298947 10.1109/TPAMI.2009.190 10.1145/1991996.1992003 10.1145/276698.276876 10.1109/ICCV.2009.5459466 10.1145/997817.997857 10.1109/CVPR.2011.5995432 10.1017/CBO9780511806292 10.1109/CVPR.2011.5995709 10.1109/TCYB.2015.2474742 10.1109/CVPR.2013.174 10.1109/TMM.2015.2419973 10.1109/TPAMI.2010.57 10.1109/ICCV.2013.377 10.1145/2502081.2502100 10.1109/CVPR.2013.388 10.1109/72.977291 10.1109/TCYB.2013.2281366 10.1016/j.patcog.2013.08.022 10.1145/2072298.2072354 10.1109/TIP.2016.2593344 10.1109/CVPR.2011.5995518 10.1007/s13735-012-0003-7 10.1145/2911451.2911502 10.1109/CVPR.2013.206 10.1145/1835804.1835946 10.1137/1.9781611973440.20 10.1109/CVPR.2010.5539994 10.1145/1646396.1646452 10.1109/CVPR.2013.378 10.1109/TPAMI.2012.48
ContentType	Journal Article
DBID	97E RIA RIE AAYXX CITATION NPM 7X8
DOI	10.1109/TIP.2017.2695895
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005-present IEEE All-Society Periodicals Package (ASPP) 1998-Present IEEE Electronic Library (IEL) CrossRef PubMed MEDLINE - Academic
DatabaseTitle	CrossRef PubMed MEDLINE - Academic
DatabaseTitleList	PubMed MEDLINE - Academic
Database_xml	– sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher – sequence: 3 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database
DeliveryMethod	fulltext_linktorsrc
Discipline	Applied Sciences Engineering
EISSN	1941-0042
EndPage	5380
ExternalDocumentID	28436872 10_1109_TIP_2017_2695895 7904662
Genre	orig-research Journal Article
GrantInformation_xml	– fundername: Foundation of State Key Laboratory of Software Development Environment grantid: 2015ZX-04 – fundername: National Natural Science Foundation of China grantid: 61402026 funderid: 10.13039/501100001809
GroupedDBID	--- -~X .DC 0R~ 29I 4.4 53G 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABFSI ABQJQ ABVLG ACGFO ACGFS ACIWK AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 E.L EBS EJD F5P HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P RIA RIE RNS TAE TN5 VH1 AAYXX CITATION AAYOK NPM RIG 7X8
ID	FETCH-LOGICAL-c319t-c1d15cf4e5849ec0f5fe38a4b594a141ad2f8208a4b50a19bd0fc9e81acd6123
IEDL.DBID	RIE
ISICitedReferencesCount	35
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000407969200021&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1057-7149 1941-0042
IngestDate	Sat Sep 27 19:52:55 EDT 2025 Thu Apr 03 07:05:31 EDT 2025 Tue Nov 18 21:18:39 EST 2025 Sat Nov 29 03:21:06 EST 2025 Tue Aug 26 17:01:01 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	11
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c319t-c1d15cf4e5849ec0f5fe38a4b594a141ad2f8208a4b50a19bd0fc9e81acd6123
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23
ORCID	0000-0001-8425-4195
PMID	28436872
PQID	1891458476
PQPubID	23479
PageCount	14
ParticipantIDs	ieee_primary_7904662 crossref_primary_10_1109_TIP_2017_2695895 pubmed_primary_28436872 proquest_miscellaneous_1891458476 crossref_citationtrail_10_1109_TIP_2017_2695895
PublicationCentury	2000
PublicationDate	2017-Nov. 2017-11-00 2017-Nov 20171101
PublicationDateYYYYMMDD	2017-11-01
PublicationDate_xml	– month: 11 year: 2017 text: 2017-Nov.
PublicationDecade	2010
PublicationPlace	United States
PublicationPlace_xml	– name: United States
PublicationTitle	IEEE transactions on image processing
PublicationTitleAbbrev	TIP
PublicationTitleAlternate	IEEE Trans Image Process
PublicationYear	2017
Publisher	IEEE
Publisher_xml	– name: IEEE
References	ref13 ref56 ref59 kong (ref62) 2012 ref58 ref14 ref53 ref52 liu (ref4) 2014 ref54 ref10 he (ref34) 2012 xia (ref21) 2015 pelillo (ref57) 2009 ref19 liu (ref51) 2017 liu (ref11) 2017 kulis (ref16) 2009 wang (ref39) 2015 ref45 ref47 ref42 ref44 erin liong (ref41) 2015 liu (ref18) 2011 heo (ref28) 2012 liu (ref55) 2010 ref8 ref3 ref6 ref5 ref40 liu (ref20) 2014 ref35 ref37 ref36 ref31 ref30 ref33 norouzi (ref43) 2011 lee (ref38) 2012 ref32 he (ref2) 2012 ref1 liu (ref26) 2012 goodman (ref12) 2004 weiss (ref15) 2008 wang (ref17) 2010 gong (ref7) 2015 sun (ref50) 2010; 32 ref24 he (ref49) 2005 ref23 ref25 ref64 ref63 ref22 ref65 ref27 wang (ref9) 2016 ref29 wang (ref46) 2016 guyon (ref48) 2003; 3 ref60 ref61
References_xml	– start-page: 1 year: 2011 ident: ref18 article-title: Hashing with graphs publication-title: Proc ICML – ident: ref5 doi: 10.1109/TCYB.2013.2289351 – ident: ref10 doi: 10.1109/TIP.2015.2505180 – ident: ref58 doi: 10.1109/CVPR.2006.130 – ident: ref31 doi: 10.1145/2540990 – ident: ref54 doi: 10.1109/TPAMI.2007.250608 – ident: ref1 doi: 10.1145/361002.361007 – start-page: 2475 year: 2015 ident: ref41 article-title: Deep hashing for compact binary codes learning publication-title: Proc IEEE Conf Comput Vis Pattern Recognit (CVPR) – ident: ref6 doi: 10.1109/CVPR.2015.7298598 – start-page: 2957 year: 2012 ident: ref28 article-title: Spherical hashing publication-title: Proc CVPR – ident: ref32 doi: 10.1016/j.neucom.2014.09.033 – start-page: 1127 year: 2012 ident: ref34 article-title: On the difficulty of nearest neighbor search publication-title: Proc ICML – ident: ref40 doi: 10.1109/CVPR.2015.7298947 – start-page: 5 year: 2012 ident: ref62 article-title: Double-bit quantization for hashing publication-title: Proc AAAI – volume: 32 start-page: 1610 year: 2010 ident: ref50 article-title: Local-learning-based feature selection for high-dimensional data analysis publication-title: IEEE Trans Pattern Anal Mach Intell doi: 10.1109/TPAMI.2009.190 – start-page: 671 year: 2010 ident: ref55 article-title: Robust graph mode seeking by graph shift publication-title: Proc ICML – start-page: 1753 year: 2008 ident: ref15 article-title: Spectral hashing publication-title: Proc NIPS – ident: ref60 doi: 10.1145/1991996.1992003 – start-page: 1127 year: 2010 ident: ref17 article-title: Sequential projection learning for hashing with compact codes publication-title: Proc ICML – ident: ref13 doi: 10.1145/276698.276876 – ident: ref27 doi: 10.1109/ICCV.2009.5459466 – start-page: 1042 year: 2009 ident: ref16 article-title: Learning to hash with binary reconstructive embeddings publication-title: Proc NIPS – ident: ref14 doi: 10.1145/997817.997857 – start-page: 1 year: 2005 ident: ref49 publication-title: Laplacian score for feature selection – ident: ref24 doi: 10.1109/CVPR.2011.5995432 – ident: ref56 doi: 10.1017/CBO9780511806292 – ident: ref61 doi: 10.1109/CVPR.2011.5995709 – ident: ref33 doi: 10.1109/TCYB.2015.2474742 – ident: ref59 doi: 10.1109/CVPR.2013.174 – ident: ref8 doi: 10.1109/TMM.2015.2419973 – ident: ref63 doi: 10.1109/TPAMI.2010.57 – ident: ref45 doi: 10.1109/ICCV.2013.377 – ident: ref44 doi: 10.1145/2502081.2502100 – ident: ref25 doi: 10.1109/CVPR.2013.388 – start-page: 1102 year: 2016 ident: ref46 article-title: Affinity preserving quantization for hashing: A vector quantization approach to learning compact binary codes publication-title: Proc AAAI – ident: ref35 doi: 10.1109/72.977291 – ident: ref3 doi: 10.1109/TCYB.2013.2281366 – start-page: 2074 year: 2012 ident: ref26 article-title: Supervised hashing with kernels publication-title: Proc CVPR – start-page: 4183 year: 2017 ident: ref51 article-title: Boosting complementary hash tables for fast nearest neighbor search publication-title: Proc AAAI – start-page: 2298 year: 2015 ident: ref39 article-title: Hamming compatible quantization for hashing publication-title: Proc AAAI – start-page: 1 year: 2014 ident: ref4 article-title: Collaborative hashin publication-title: Proc IEEE CVPR – ident: ref30 doi: 10.1016/j.patcog.2013.08.022 – start-page: 712 year: 2009 ident: ref57 article-title: What is a cluster? perspectives from game theory publication-title: Proc of the PASCAL – ident: ref29 doi: 10.1145/2072298.2072354 – start-page: 3332 year: 2015 ident: ref21 article-title: Sparse projections for high-dimensional binary codes publication-title: Proc IEEE CVPR – year: 2004 ident: ref12 publication-title: Nearest Neighbors High-Dimensional Spaces Handbook Discrete Computer Geometry – start-page: 3419 year: 2014 ident: ref20 article-title: Discrete graph hashing publication-title: Proc NIPS – ident: ref22 doi: 10.1109/TIP.2016.2593344 – ident: ref23 doi: 10.1109/CVPR.2011.5995518 – start-page: 214 year: 2012 ident: ref38 article-title: Quadra-embedding: Binary code embedding with low quantization error publication-title: Proc ACCV – start-page: 2181 year: 2016 ident: ref9 article-title: To project more or to quantize more: Minimize reconstruction bias for learning compact binary codes publication-title: Proc IJCAI – start-page: 1 year: 2017 ident: ref11 article-title: Deep sketch hashing: Fast free-hand sketch-based image retrieval publication-title: Proc IEEE CVPR – ident: ref47 doi: 10.1007/s13735-012-0003-7 – ident: ref37 doi: 10.1145/2911451.2911502 – ident: ref36 doi: 10.1109/CVPR.2013.206 – ident: ref52 doi: 10.1145/1835804.1835946 – volume: 3 start-page: 1157 year: 2003 ident: ref48 article-title: An introduction to variable and feature selection publication-title: J Mach Learn Res – ident: ref65 doi: 10.1137/1.9781611973440.20 – ident: ref42 doi: 10.1109/CVPR.2010.5539994 – start-page: 19 year: 2015 ident: ref7 article-title: Web scale photo hash clustering on a single machine publication-title: Proc IEEE CVPR – ident: ref64 doi: 10.1145/1646396.1646452 – start-page: 3005 year: 2012 ident: ref2 article-title: Mobile product search with bag of hash bits and boundary reranking publication-title: Proc CVPR – ident: ref19 doi: 10.1109/CVPR.2013.378 – start-page: 353 year: 2011 ident: ref43 article-title: Minimal loss hashing for compact binary codes publication-title: Proc ICML – ident: ref53 doi: 10.1109/TPAMI.2012.48
SSID	ssj0014516
Score	2.4192731
Snippet	To overcome the barrier of storage and computation when dealing with gigantic-scale data sets, compact hashing has been studied extensively to approximate the...
SourceID	proquest pubmed crossref ieee
SourceType	Aggregation Database Index Database Enrichment Source Publisher
StartPage	5367
SubjectTerms	Algorithm design and analysis Binary codes bit complementarity bit reliability Dynamic programming hash bit selection Heuristic algorithms locality-sensitive hashing Nearest neighbor search Nearest neighbor searches normalized dominant set Optimization Reliability
Title	Hash Bit Selection for Nearest Neighbor Search
URI	https://ieeexplore.ieee.org/document/7904662 https://www.ncbi.nlm.nih.gov/pubmed/28436872 https://www.proquest.com/docview/1891458476
Volume	26
WOSCitedRecordID	wos000407969200021&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
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1941-0042 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014516 issn: 1057-7149 databaseCode: RIE dateStart: 19920101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1bS8MwFD7M4YM-ON281Muo4Itgt2Zrm-RRxTFBxsAheytpkuJANlk7f78n6QUFFXxqKUlb8p2TfqcnOR_AlVKBiIY08ZRKqBdIDFCYJKnHRDDUxgioVYl4eaKTCZvP-bQBN_VeGK21XXyme-bU5vLVSm7Mr7I-5RjNmQl3i1Ja7NWqMwZGcNZmNkPqUaT9VUrS5_3Z49Ss4aK9QcRDxo1YDU7Kw4jRwbevkZVX-Z1p2i_OqPW_d92HvZJZureFKRxAQy_b0CpZplv6cNaG3S8lCDvQG4vs1b1b5O6zVcRBmFzkse7E1LbNcjxi9I524hbrkg9hNnqY3Y-9UkPBk-hcuSeJIqFMA41Eg2vpp2GqhwhEEvJAkIAINUiRBNgrviA8UX4quWZESGUqsxxBc7la6hNwGfp2wAdEGg7FI5IInooEu2oknZFPHehXQxnLsr64kbl4i22c4fMYcYgNDnGJgwPXdY_3orbGH207ZozrduXwOnBZoRWjX5hkh1jq1SaLCePE5oAjB44LGOvOFfqnP9_0DHbMo4sdh-fQzNcbfQHb8iNfZOsuGt-cda3xfQK21NCk
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1bS8MwFD4MFdQH75d6reCLYLemS5vmUcWx4RwDh_hW0iRFQTaxnb_fk_SCggo-tZSklHznJN_pSc4HcK4UFVGXpZ5SKfOoxAAlliTzYkG72hgBsyoRj0M2GsVPT3zcgsvmLIzW2m4-021za3P5aibn5ldZh3GM5syEuxhSGpDytFaTMzCSsza3GTKPIfGvk5I-70wGY7OLi7WDiIcxN3I1OC13o5gF39YjK7DyO9e0a05v_X9fuwFrFbd0r0pj2ISWnm7BesUz3cqL8y1Y_VKEcBvafZE_u9cvhftgNXEQKBeZrDsy1W3zAq8Yv6OluOXO5B2Y9G4nN32vUlHwJLpX4UmiSCgzqpFqcC39LMx0F6FIQ04FoUSoIEMaYJ_4gvBU-ZnkOiZCKlObZRcWprOp3gc3Ru-mPCDSsCgekVTwTKTYVSPtjHzmQKceykRWFcaN0MVrYiMNnyeIQ2JwSCocHLhoeryV1TX-aLttxrhpVw2vA2c1Wgl6hkl3iKmezfOExJzYLHDkwF4JY9O5Rv_g55eewnJ_cj9MhoPR3SGsmM8ozx8ewULxPtfHsCQ_ipf8_cSa4CfTRNMD
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%3Ajournal&rft.genre=article&rft.atitle=Hash+Bit+Selection+for+Nearest+Neighbor+Search&rft.jtitle=IEEE+transactions+on+image+processing&rft.au=Liu%2C+Xianglong&rft.au=He%2C+Junfeng&rft.au=Chang%2C+Shih-Fu&rft.date=2017-11-01&rft.issn=1057-7149&rft.eissn=1941-0042&rft.volume=26&rft.issue=11&rft.spage=5367&rft.epage=5380&rft_id=info:doi/10.1109%2FTIP.2017.2695895&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TIP_2017_2695895
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1057-7149&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1057-7149&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1057-7149&client=summon