Nonlinear Discrete Hashing

In this paper, we propose a nonlinear discrete hashing approach to learn compact binary codes for scalable image search. Instead of seeking a single linear projection in most existing hashing methods, we pursue a multilayer network with nonlinear transformations to capture the local structure of dat...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on multimedia 2017-01, Vol.19 (1), p.123-135
Main Authors: Chen, Zhixiang, Lu, Jiwen, Feng, Jianjiang, Zhou, Jie
Format: Article
Language:English
Subjects:
Binary code
Binary codes
Data structures
discrete optimization
Errors
hashing
Manuals
Multi-layer neural network
multilayer neural network
nonlinear transformation
Nonlinearity
Optimization
Quantization (signal)
Semantics
Similarity
Transformations
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713
cites cdi_FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713
container_end_page 135
container_issue 1
container_start_page 123
container_title IEEE transactions on multimedia
container_volume 19
creator Chen, Zhixiang
Lu, Jiwen
Feng, Jianjiang
Zhou, Jie
description In this paper, we propose a nonlinear discrete hashing approach to learn compact binary codes for scalable image search. Instead of seeking a single linear projection in most existing hashing methods, we pursue a multilayer network with nonlinear transformations to capture the local structure of data samples. Unlike most existing hashing methods that adopt an error-prone relaxation to learn the transformations, we directly solve the discrete optimization problem to eliminate the quantization error accumulation. Specifically, to leverage the similarity relationships between data samples and exploit the semantic affinities of manual labels, the binary codes are learned with the objective to: 1) minimize the quantization error between the original data samples and the learned binary codes; 2) preserve the similarity relationships in the learned binary codes; 3) maximize the information content with independent bits; and 4) maximize the accuracy of the predicted labels based on the binary codes. With an alternating optimization, the nonlinear transformation and the discrete quantization are jointly optimized in the hashing learning framework. Experimental results on four datasets including CIFAR10, MNIST, SUN397, and ILSVRC2012 demonstrate that the proposed approach is superior to several state-of-the-art hashing methods.
doi_str_mv 10.1109/TMM.2016.2620604
format article
fullrecord <record><control><sourceid>proquest_ieee_</sourceid><recordid>TN_cdi_proquest_journals_1850211026</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>7637026</ieee_id><sourcerecordid>1850211026</sourcerecordid><originalsourceid>FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713</originalsourceid><addsrcrecordid>eNo9kD1PwzAQhi0EEqWwI1gqMafcnRO7HlH5KFILS5ktx7lAqpIUOx349zhqxXTv8H7oHiGuEaaIYO7Xq9WUANWUFIGC_ESM0OSYAWh9mnRBkBlCOBcXMW4AMC9Aj8TNW9dum5ZdmDw20QfuebJw8atpPy_FWe22ka-Odyw-np_W80W2fH95nT8sM08G-0yWXpuasTQlsXa-AiorNBKqOilpdKkMgZtJZySRL9jpwnknDVSKlEY5FneH3l3ofvYce7vp9qFNkxZnBVB6j1RywcHlQxdj4NruQvPtwq9FsAMBmwjYgYA9EkiR20OkYeZ_u1ZSD4V__6VUlQ</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>1850211026</pqid></control><display><type>article</type><title>Nonlinear Discrete Hashing</title><source>IEEE Electronic Library (IEL) Journals</source><creator>Chen, Zhixiang ; Lu, Jiwen ; Feng, Jianjiang ; Zhou, Jie</creator><creatorcontrib>Chen, Zhixiang ; Lu, Jiwen ; Feng, Jianjiang ; Zhou, Jie</creatorcontrib><description>In this paper, we propose a nonlinear discrete hashing approach to learn compact binary codes for scalable image search. Instead of seeking a single linear projection in most existing hashing methods, we pursue a multilayer network with nonlinear transformations to capture the local structure of data samples. Unlike most existing hashing methods that adopt an error-prone relaxation to learn the transformations, we directly solve the discrete optimization problem to eliminate the quantization error accumulation. Specifically, to leverage the similarity relationships between data samples and exploit the semantic affinities of manual labels, the binary codes are learned with the objective to: 1) minimize the quantization error between the original data samples and the learned binary codes; 2) preserve the similarity relationships in the learned binary codes; 3) maximize the information content with independent bits; and 4) maximize the accuracy of the predicted labels based on the binary codes. With an alternating optimization, the nonlinear transformation and the discrete quantization are jointly optimized in the hashing learning framework. Experimental results on four datasets including CIFAR10, MNIST, SUN397, and ILSVRC2012 demonstrate that the proposed approach is superior to several state-of-the-art hashing methods.</description><identifier>ISSN: 1520-9210</identifier><identifier>EISSN: 1941-0077</identifier><identifier>DOI: 10.1109/TMM.2016.2620604</identifier><identifier>CODEN: ITMUF8</identifier><language>eng</language><publisher>Piscataway: IEEE</publisher><subject>Binary code ; Binary codes ; Data structures ; discrete optimization ; Errors ; hashing ; Manuals ; Multi-layer neural network ; multilayer neural network ; nonlinear transformation ; Nonlinearity ; Optimization ; Quantization (signal) ; Semantics ; Similarity ; Transformations</subject><ispartof>IEEE transactions on multimedia, 2017-01, Vol.19 (1), p.123-135</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2017</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713</citedby><cites>FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/7637026$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>315,786,790,27957,27958,55147</link.rule.ids></links><search><creatorcontrib>Chen, Zhixiang</creatorcontrib><creatorcontrib>Lu, Jiwen</creatorcontrib><creatorcontrib>Feng, Jianjiang</creatorcontrib><creatorcontrib>Zhou, Jie</creatorcontrib><title>Nonlinear Discrete Hashing</title><title>IEEE transactions on multimedia</title><addtitle>TMM</addtitle><description>In this paper, we propose a nonlinear discrete hashing approach to learn compact binary codes for scalable image search. Instead of seeking a single linear projection in most existing hashing methods, we pursue a multilayer network with nonlinear transformations to capture the local structure of data samples. Unlike most existing hashing methods that adopt an error-prone relaxation to learn the transformations, we directly solve the discrete optimization problem to eliminate the quantization error accumulation. Specifically, to leverage the similarity relationships between data samples and exploit the semantic affinities of manual labels, the binary codes are learned with the objective to: 1) minimize the quantization error between the original data samples and the learned binary codes; 2) preserve the similarity relationships in the learned binary codes; 3) maximize the information content with independent bits; and 4) maximize the accuracy of the predicted labels based on the binary codes. With an alternating optimization, the nonlinear transformation and the discrete quantization are jointly optimized in the hashing learning framework. Experimental results on four datasets including CIFAR10, MNIST, SUN397, and ILSVRC2012 demonstrate that the proposed approach is superior to several state-of-the-art hashing methods.</description><subject>Binary code</subject><subject>Binary codes</subject><subject>Data structures</subject><subject>discrete optimization</subject><subject>Errors</subject><subject>hashing</subject><subject>Manuals</subject><subject>Multi-layer neural network</subject><subject>multilayer neural network</subject><subject>nonlinear transformation</subject><subject>Nonlinearity</subject><subject>Optimization</subject><subject>Quantization (signal)</subject><subject>Semantics</subject><subject>Similarity</subject><subject>Transformations</subject><issn>1520-9210</issn><issn>1941-0077</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2017</creationdate><recordtype>article</recordtype><recordid>eNo9kD1PwzAQhi0EEqWwI1gqMafcnRO7HlH5KFILS5ktx7lAqpIUOx349zhqxXTv8H7oHiGuEaaIYO7Xq9WUANWUFIGC_ESM0OSYAWh9mnRBkBlCOBcXMW4AMC9Aj8TNW9dum5ZdmDw20QfuebJw8atpPy_FWe22ka-Odyw-np_W80W2fH95nT8sM08G-0yWXpuasTQlsXa-AiorNBKqOilpdKkMgZtJZySRL9jpwnknDVSKlEY5FneH3l3ofvYce7vp9qFNkxZnBVB6j1RywcHlQxdj4NruQvPtwq9FsAMBmwjYgYA9EkiR20OkYeZ_u1ZSD4V__6VUlQ</recordid><startdate>201701</startdate><enddate>201701</enddate><creator>Chen, Zhixiang</creator><creator>Lu, Jiwen</creator><creator>Feng, Jianjiang</creator><creator>Zhou, Jie</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope></search><sort><creationdate>201701</creationdate><title>Nonlinear Discrete Hashing</title><author>Chen, Zhixiang ; Lu, Jiwen ; Feng, Jianjiang ; Zhou, Jie</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2017</creationdate><topic>Binary code</topic><topic>Binary codes</topic><topic>Data structures</topic><topic>discrete optimization</topic><topic>Errors</topic><topic>hashing</topic><topic>Manuals</topic><topic>Multi-layer neural network</topic><topic>multilayer neural network</topic><topic>nonlinear transformation</topic><topic>Nonlinearity</topic><topic>Optimization</topic><topic>Quantization (signal)</topic><topic>Semantics</topic><topic>Similarity</topic><topic>Transformations</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Chen, Zhixiang</creatorcontrib><creatorcontrib>Lu, Jiwen</creatorcontrib><creatorcontrib>Feng, Jianjiang</creatorcontrib><creatorcontrib>Zhou, Jie</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Xplore</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>IEEE transactions on multimedia</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Chen, Zhixiang</au><au>Lu, Jiwen</au><au>Feng, Jianjiang</au><au>Zhou, Jie</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Nonlinear Discrete Hashing</atitle><jtitle>IEEE transactions on multimedia</jtitle><stitle>TMM</stitle><date>2017-01</date><risdate>2017</risdate><volume>19</volume><issue>1</issue><spage>123</spage><epage>135</epage><pages>123-135</pages><issn>1520-9210</issn><eissn>1941-0077</eissn><coden>ITMUF8</coden><abstract>In this paper, we propose a nonlinear discrete hashing approach to learn compact binary codes for scalable image search. Instead of seeking a single linear projection in most existing hashing methods, we pursue a multilayer network with nonlinear transformations to capture the local structure of data samples. Unlike most existing hashing methods that adopt an error-prone relaxation to learn the transformations, we directly solve the discrete optimization problem to eliminate the quantization error accumulation. Specifically, to leverage the similarity relationships between data samples and exploit the semantic affinities of manual labels, the binary codes are learned with the objective to: 1) minimize the quantization error between the original data samples and the learned binary codes; 2) preserve the similarity relationships in the learned binary codes; 3) maximize the information content with independent bits; and 4) maximize the accuracy of the predicted labels based on the binary codes. With an alternating optimization, the nonlinear transformation and the discrete quantization are jointly optimized in the hashing learning framework. Experimental results on four datasets including CIFAR10, MNIST, SUN397, and ILSVRC2012 demonstrate that the proposed approach is superior to several state-of-the-art hashing methods.</abstract><cop>Piscataway</cop><pub>IEEE</pub><doi>10.1109/TMM.2016.2620604</doi><tpages>13</tpages></addata></record>
fulltext fulltext
identifier ISSN: 1520-9210
ispartof IEEE transactions on multimedia, 2017-01, Vol.19 (1), p.123-135
issn 1520-9210
1941-0077
language eng
recordid cdi_proquest_journals_1850211026
source IEEE Electronic Library (IEL) Journals
subjects Binary code
Binary codes
Data structures
discrete optimization
Errors
hashing
Manuals
Multi-layer neural network
multilayer neural network
nonlinear transformation
Nonlinearity
Optimization
Quantization (signal)
Semantics
Similarity
Transformations
title Nonlinear Discrete Hashing
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-09-23T04%3A32%3A06IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_ieee_&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Nonlinear%20Discrete%20Hashing&rft.jtitle=IEEE%20transactions%20on%20multimedia&rft.au=Chen,%20Zhixiang&rft.date=2017-01&rft.volume=19&rft.issue=1&rft.spage=123&rft.epage=135&rft.pages=123-135&rft.issn=1520-9210&rft.eissn=1941-0077&rft.coden=ITMUF8&rft_id=info:doi/10.1109/TMM.2016.2620604&rft_dat=%3Cproquest_ieee_%3E1850211026%3C/proquest_ieee_%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c291t-3bc79fe1b9b2e7acd02bd1930df02b397b6920a83a9322c5ea75aca390d626713%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=1850211026&rft_id=info:pmid/&rft_ieee_id=7637026&rfr_iscdi=true