{"id":"https://openalex.org/W2963404959","doi":"https://doi.org/10.1109/ijcnn.2018.8489312","title":"Diagonalwise Refactorization: An Efficient Training Method for Depthwise Convolutions","display_name":"Diagonalwise Refactorization: An Efficient Training Method for Depthwise Convolutions","publication_year":2018,"publication_date":"2018-07-01","ids":{"openalex":"https://openalex.org/W2963404959","doi":"https://doi.org/10.1109/ijcnn.2018.8489312","mag":"2963404959"},"language":"en","primary_location":{"id":"doi:10.1109/ijcnn.2018.8489312","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ijcnn.2018.8489312","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2018 International Joint Conference on Neural Networks (IJCNN)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100695329","display_name":"Zheng Qin","orcid":"https://orcid.org/0000-0003-4373-4111"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Zheng Qin","raw_affiliation_strings":["Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China"],"affiliations":[{"raw_affiliation_string":"Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050642628","display_name":"Zhaoning Zhang","orcid":"https://orcid.org/0000-0001-7518-1385"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhaoning Zhang","raw_affiliation_strings":["Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China"],"affiliations":[{"raw_affiliation_string":"Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100440903","display_name":"Dongsheng Li","orcid":"https://orcid.org/0000-0001-9743-2034"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dongsheng Li","raw_affiliation_strings":["Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China"],"affiliations":[{"raw_affiliation_string":"Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100395351","display_name":"Yiming Zhang","orcid":"https://orcid.org/0000-0001-6450-8485"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yiming Zhang","raw_affiliation_strings":["Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China"],"affiliations":[{"raw_affiliation_string":"Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101645557","display_name":"Yuxing Peng","orcid":"https://orcid.org/0000-0002-8223-5588"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuxing Peng","raw_affiliation_strings":["Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China"],"affiliations":[{"raw_affiliation_string":"Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5100695329"],"corresponding_institution_ids":["https://openalex.org/I170215575"],"apc_list":null,"apc_paid":null,"fwci":1.2535,"has_fulltext":false,"cited_by_count":28,"citation_normalized_percentile":{"value":0.85427549,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"8"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10812","display_name":"Human Pose and Action Recognition","score":0.9987999796867371,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9987000226974487,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8562690019607544},{"id":"https://openalex.org/keywords/convolution","display_name":"Convolution (computer science)","score":0.7869383096694946},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6947119832038879},{"id":"https://openalex.org/keywords/diagonal","display_name":"Diagonal","score":0.5663087368011475},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5591715574264526},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.5152173638343811},{"id":"https://openalex.org/keywords/implementation","display_name":"Implementation","score":0.46923667192459106},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.43088293075561523},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.41413113474845886},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4090243875980377},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3857976198196411},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.3418717086315155},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.1945629119873047},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10009685158729553},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.07603448629379272}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8562690019607544},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.7869383096694946},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6947119832038879},{"id":"https://openalex.org/C130367717","wikidata":"https://www.wikidata.org/wiki/Q189791","display_name":"Diagonal","level":2,"score":0.5663087368011475},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5591715574264526},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.5152173638343811},{"id":"https://openalex.org/C26713055","wikidata":"https://www.wikidata.org/wiki/Q245962","display_name":"Implementation","level":2,"score":0.46923667192459106},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.43088293075561523},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.41413113474845886},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4090243875980377},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3857976198196411},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.3418717086315155},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.1945629119873047},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10009685158729553},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.07603448629379272},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ijcnn.2018.8489312","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ijcnn.2018.8489312","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2018 International Joint Conference on Neural Networks (IJCNN)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":64,"referenced_works":["https://openalex.org/W639708223","https://openalex.org/W753012316","https://openalex.org/W1667652561","https://openalex.org/W1686810756","https://openalex.org/W1902934009","https://openalex.org/W2097117768","https://openalex.org/W2104636679","https://openalex.org/W2119144962","https://openalex.org/W2155893237","https://openalex.org/W2163605009","https://openalex.org/W2167215970","https://openalex.org/W2186615578","https://openalex.org/W2194775991","https://openalex.org/W2267635276","https://openalex.org/W2279098554","https://openalex.org/W2300242332","https://openalex.org/W2402144811","https://openalex.org/W2469490737","https://openalex.org/W2531409750","https://openalex.org/W2549139847","https://openalex.org/W2557728737","https://openalex.org/W2570343428","https://openalex.org/W2593245696","https://openalex.org/W2611105550","https://openalex.org/W2612445135","https://openalex.org/W2613718673","https://openalex.org/W2950248853","https://openalex.org/W2953384591","https://openalex.org/W2962851801","https://openalex.org/W2963048316","https://openalex.org/W2963114950","https://openalex.org/W2963125010","https://openalex.org/W2963363373","https://openalex.org/W2963674932","https://openalex.org/W2963844898","https://openalex.org/W2963881378","https://openalex.org/W2964228333","https://openalex.org/W2964299589","https://openalex.org/W4295262505","https://openalex.org/W4297775537","https://openalex.org/W4302296459","https://openalex.org/W6620707391","https://openalex.org/W6622239757","https://openalex.org/W6637151318","https://openalex.org/W6637373629","https://openalex.org/W6639703010","https://openalex.org/W6677580257","https://openalex.org/W6679667936","https://openalex.org/W6684191040","https://openalex.org/W6684563725","https://openalex.org/W6686509673","https://openalex.org/W6693397755","https://openalex.org/W6695314431","https://openalex.org/W6698200048","https://openalex.org/W6713134421","https://openalex.org/W6720242923","https://openalex.org/W6728184133","https://openalex.org/W6729342207","https://openalex.org/W6730028961","https://openalex.org/W6731892127","https://openalex.org/W6734062232","https://openalex.org/W6737664043","https://openalex.org/W6740474667","https://openalex.org/W6746264161"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W2146343568","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W2013643406","https://openalex.org/W2027972911","https://openalex.org/W2157978810","https://openalex.org/W2597809628","https://openalex.org/W3157543420"],"abstract_inverted_index":{"Depthwise":[0],"convolutions":[1],"provide":[2],"significant":[3],"performance":[4],"benefits":[5],"owing":[6],"to":[7,65,82,145],"the":[8,35,55,67,84,97],"reduction":[9],"in":[10,24,42,112],"both":[11],"parameters":[12],"and":[13,140],"mult-adds.":[14],"However,":[15],"training":[16,56,110,127],"depthwise":[17,58,72,85,150],"convolution":[18,59,73,86],"layers":[19],"with":[20],"GPUs":[21],"is":[22,64,93,101],"slow":[23],"current":[25],"deep":[26,115],"learning":[27,116],"frameworks":[28],"because":[29],"their":[30,146],"implementations":[31,148],"cannot":[32],"fully":[33],"utilize":[34],"GPU":[36,104],"capacity.":[37],"To":[38],"address":[39],"this":[40,43],"problem,":[41],"paper":[44],"we":[45],"present":[46],"an":[47],"efficient":[48],"method":[49,111,124],"(called":[50],"diagonalwise":[51],"refactorization)":[52],"for":[53,103],"accelerating":[54],"of":[57,70,149],"layers.":[60],"Our":[61],"key":[62],"idea":[63],"rearrange":[66],"weight":[68,78],"vectors":[69],"a":[71,75],"into":[74,87],"large":[76],"diagonal":[77],"matrix":[79],"so":[80],"as":[81],"convert":[83],"one":[88],"single":[89],"standard":[90],"convolution,":[91],"which":[92],"well":[94],"supported":[95],"by":[96],"cuDNN":[98],"library":[99],"that":[100,121],"highly-optimized":[102],"computations.":[105],"We":[106],"have":[107],"implemented":[108],"our":[109,122],"five":[113],"popular":[114],"frameworks.":[117],"Evaluation":[118],"results":[119],"show":[120],"proposed":[123],"gains":[125],"15.4\u00d7":[126],"speedup":[128],"on":[129,132,135,138,142],"Darknet,":[130],"8.4\u00d7":[131],"Caffe,":[133],"5.4\u00d7":[134],"PyTorch,":[136],"3.5\u00d7":[137],"MXNet,":[139],"1.4\u00d7":[141],"TensorFlow,":[143],"compared":[144],"original":[147],"convolutions.":[151]},"counts_by_year":[{"year":2024,"cited_by_count":6},{"year":2023,"cited_by_count":6},{"year":2022,"cited_by_count":4},{"year":2021,"cited_by_count":3},{"year":2020,"cited_by_count":6},{"year":2019,"cited_by_count":3}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}