Fast Inference of Binarized Convolutional Neural Networks Exploiting Max Pooling with Modified Block Structure
This letter presents a novel technique to achieve a fast inference of the binarized convolutional neural networks (BCNN). The proposed technique modifies the structure of the constituent blocks of the BCNN model so that the input elements for the max-pooling operation are binary. In this structure,...
Gespeichert in:
| Veröffentlicht in: | IEICE Transactions on Information and Systems 2020/03/01, Vol.E103.D(3), pp.706-710 |
|---|---|
| Hauptverfasser: | , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 710 |
|---|---|
| container_issue | 3 |
| container_start_page | 706 |
| container_title | IEICE Transactions on Information and Systems |
| container_volume | E103.D |
| creator | SHIN, Ji-Hoon KIM, Tae-Hwan |
| description | This letter presents a novel technique to achieve a fast inference of the binarized convolutional neural networks (BCNN). The proposed technique modifies the structure of the constituent blocks of the BCNN model so that the input elements for the max-pooling operation are binary. In this structure, if any of the input elements is +1, the result of the pooling can be produced immediately; the proposed technique eliminates such computations that are involved to obtain the remaining input elements, so as to reduce the inference time effectively. The proposed technique reduces the inference time by up to 34.11%, while maintaining the classification accuracy. |
| doi_str_mv | 10.1587/transinf.2019EDL8165 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_journals_2369530579</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2369530579</sourcerecordid><originalsourceid>FETCH-LOGICAL-c466t-ded17ef48e96118f442807a06e198d5f7f2ac066f82750151d198f877ff9ffa13</originalsourceid><addsrcrecordid>eNpNkElPwzAUhC0EEmX5BxwscQ74JfGSIy0tIJVFLGfLSmxwCXaxHVr49QTK0tOMnuYb6Q1CB0COgAp-nIJy0TpzlBOoxqdTAYxuoAHwkmZQMNhEA1IBywQt8m20E-OMEBA50AFyExUTvnBGB-1qjb3BQ-tUsB-6wSPv3nzbJeudavGV7sK3pIUPzxGPl_PW22TdI75US3zjffvlFzY94UvfWGP7imHr62d8l0JXpy7oPbRlVBv1_o_uoofJ-H50nk2vzy5GJ9OsLhlLWaMb4NqUQlcMQJiyzAXhijANlWio4SZXNWHMiJxTAhSa_m4E58ZUxigodtHhqnce_GunY5Iz34X-iyjzglW0IJRXfapcpergYwzayHmwLyq8SyDya1n5u6xcW7bHblfYLCb1qP8gFZKtW_0PjYEU8lQWv2at5C9cP6kgtSs-AS-2jOc</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2369530579</pqid></control><display><type>article</type><title>Fast Inference of Binarized Convolutional Neural Networks Exploiting Max Pooling with Modified Block Structure</title><source>J-STAGE Free</source><source>EZB-FREE-00999 freely available EZB journals</source><creator>SHIN, Ji-Hoon ; KIM, Tae-Hwan</creator><creatorcontrib>SHIN, Ji-Hoon ; KIM, Tae-Hwan</creatorcontrib><description>This letter presents a novel technique to achieve a fast inference of the binarized convolutional neural networks (BCNN). The proposed technique modifies the structure of the constituent blocks of the BCNN model so that the input elements for the max-pooling operation are binary. In this structure, if any of the input elements is +1, the result of the pooling can be produced immediately; the proposed technique eliminates such computations that are involved to obtain the remaining input elements, so as to reduce the inference time effectively. The proposed technique reduces the inference time by up to 34.11%, while maintaining the classification accuracy.</description><identifier>ISSN: 0916-8532</identifier><identifier>EISSN: 1745-1361</identifier><identifier>DOI: 10.1587/transinf.2019EDL8165</identifier><language>eng</language><publisher>Tokyo: The Institute of Electronics, Information and Communication Engineers</publisher><subject>Artificial neural networks ; binarized neural networks ; convolutional neural networks ; deep learning ; embedded systems ; Inference ; Neural networks</subject><ispartof>IEICE Transactions on Information and Systems, 2020/03/01, Vol.E103.D(3), pp.706-710</ispartof><rights>2020 The Institute of Electronics, Information and Communication Engineers</rights><rights>Copyright Japan Science and Technology Agency 2020</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c466t-ded17ef48e96118f442807a06e198d5f7f2ac066f82750151d198f877ff9ffa13</citedby><cites>FETCH-LOGICAL-c466t-ded17ef48e96118f442807a06e198d5f7f2ac066f82750151d198f877ff9ffa13</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,780,784,1883,27924,27925</link.rule.ids></links><search><creatorcontrib>SHIN, Ji-Hoon</creatorcontrib><creatorcontrib>KIM, Tae-Hwan</creatorcontrib><title>Fast Inference of Binarized Convolutional Neural Networks Exploiting Max Pooling with Modified Block Structure</title><title>IEICE Transactions on Information and Systems</title><addtitle>IEICE Trans. Inf. & Syst.</addtitle><description>This letter presents a novel technique to achieve a fast inference of the binarized convolutional neural networks (BCNN). The proposed technique modifies the structure of the constituent blocks of the BCNN model so that the input elements for the max-pooling operation are binary. In this structure, if any of the input elements is +1, the result of the pooling can be produced immediately; the proposed technique eliminates such computations that are involved to obtain the remaining input elements, so as to reduce the inference time effectively. The proposed technique reduces the inference time by up to 34.11%, while maintaining the classification accuracy.</description><subject>Artificial neural networks</subject><subject>binarized neural networks</subject><subject>convolutional neural networks</subject><subject>deep learning</subject><subject>embedded systems</subject><subject>Inference</subject><subject>Neural networks</subject><issn>0916-8532</issn><issn>1745-1361</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><recordid>eNpNkElPwzAUhC0EEmX5BxwscQ74JfGSIy0tIJVFLGfLSmxwCXaxHVr49QTK0tOMnuYb6Q1CB0COgAp-nIJy0TpzlBOoxqdTAYxuoAHwkmZQMNhEA1IBywQt8m20E-OMEBA50AFyExUTvnBGB-1qjb3BQ-tUsB-6wSPv3nzbJeudavGV7sK3pIUPzxGPl_PW22TdI75US3zjffvlFzY94UvfWGP7imHr62d8l0JXpy7oPbRlVBv1_o_uoofJ-H50nk2vzy5GJ9OsLhlLWaMb4NqUQlcMQJiyzAXhijANlWio4SZXNWHMiJxTAhSa_m4E58ZUxigodtHhqnce_GunY5Iz34X-iyjzglW0IJRXfapcpergYwzayHmwLyq8SyDya1n5u6xcW7bHblfYLCb1qP8gFZKtW_0PjYEU8lQWv2at5C9cP6kgtSs-AS-2jOc</recordid><startdate>20200301</startdate><enddate>20200301</enddate><creator>SHIN, Ji-Hoon</creator><creator>KIM, Tae-Hwan</creator><general>The Institute of Electronics, Information and Communication Engineers</general><general>Japan Science and Technology Agency</general><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope></search><sort><creationdate>20200301</creationdate><title>Fast Inference of Binarized Convolutional Neural Networks Exploiting Max Pooling with Modified Block Structure</title><author>SHIN, Ji-Hoon ; KIM, Tae-Hwan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c466t-ded17ef48e96118f442807a06e198d5f7f2ac066f82750151d198f877ff9ffa13</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Artificial neural networks</topic><topic>binarized neural networks</topic><topic>convolutional neural networks</topic><topic>deep learning</topic><topic>embedded systems</topic><topic>Inference</topic><topic>Neural networks</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>SHIN, Ji-Hoon</creatorcontrib><creatorcontrib>KIM, Tae-Hwan</creatorcontrib><collection>CrossRef</collection><collection>Computer and Information Systems 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>IEICE Transactions on Information and Systems</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>SHIN, Ji-Hoon</au><au>KIM, Tae-Hwan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Fast Inference of Binarized Convolutional Neural Networks Exploiting Max Pooling with Modified Block Structure</atitle><jtitle>IEICE Transactions on Information and Systems</jtitle><addtitle>IEICE Trans. Inf. & Syst.</addtitle><date>2020-03-01</date><risdate>2020</risdate><volume>E103.D</volume><issue>3</issue><spage>706</spage><epage>710</epage><pages>706-710</pages><issn>0916-8532</issn><eissn>1745-1361</eissn><abstract>This letter presents a novel technique to achieve a fast inference of the binarized convolutional neural networks (BCNN). The proposed technique modifies the structure of the constituent blocks of the BCNN model so that the input elements for the max-pooling operation are binary. In this structure, if any of the input elements is +1, the result of the pooling can be produced immediately; the proposed technique eliminates such computations that are involved to obtain the remaining input elements, so as to reduce the inference time effectively. The proposed technique reduces the inference time by up to 34.11%, while maintaining the classification accuracy.</abstract><cop>Tokyo</cop><pub>The Institute of Electronics, Information and Communication Engineers</pub><doi>10.1587/transinf.2019EDL8165</doi><tpages>5</tpages><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0916-8532 |
| ispartof | IEICE Transactions on Information and Systems, 2020/03/01, Vol.E103.D(3), pp.706-710 |
| issn | 0916-8532 1745-1361 |
| language | eng |
| recordid | cdi_proquest_journals_2369530579 |
| source | J-STAGE Free; EZB-FREE-00999 freely available EZB journals |
| subjects | Artificial neural networks binarized neural networks convolutional neural networks deep learning embedded systems Inference Neural networks |
| title | Fast Inference of Binarized Convolutional Neural Networks Exploiting Max Pooling with Modified Block Structure |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-18T21%3A29%3A21IST&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=Fast%20Inference%20of%20Binarized%20Convolutional%20Neural%20Networks%20Exploiting%20Max%20Pooling%20with%20Modified%20Block%20Structure&rft.jtitle=IEICE%20Transactions%20on%20Information%20and%20Systems&rft.au=SHIN,%20Ji-Hoon&rft.date=2020-03-01&rft.volume=E103.D&rft.issue=3&rft.spage=706&rft.epage=710&rft.pages=706-710&rft.issn=0916-8532&rft.eissn=1745-1361&rft_id=info:doi/10.1587/transinf.2019EDL8165&rft_dat=%3Cproquest_cross%3E2369530579%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=2369530579&rft_id=info:pmid/&rfr_iscdi=true |