{"id":"https://openalex.org/W4390658705","doi":"https://doi.org/10.1145/3639823","title":"An Efficient Hybrid Deep Learning Accelerator for Compact and Heterogeneous CNNs","display_name":"An Efficient Hybrid Deep Learning Accelerator for Compact and Heterogeneous CNNs","publication_year":2024,"publication_date":"2024-01-08","ids":{"openalex":"https://openalex.org/W4390658705","doi":"https://doi.org/10.1145/3639823"},"language":"en","primary_location":{"id":"doi:10.1145/3639823","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3639823","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3639823","source":{"id":"https://openalex.org/S26056741","display_name":"ACM Transactions on Architecture and Code Optimization","issn_l":"1544-3566","issn":["1544-3566","1544-3973"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Architecture and Code Optimization","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3639823","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5057731855","display_name":"Fareed Qararyah","orcid":"https://orcid.org/0000-0002-3955-2836"},"institutions":[{"id":"https://openalex.org/I66862912","display_name":"Chalmers University of Technology","ror":"https://ror.org/040wg7k59","country_code":"SE","type":"education","lineage":["https://openalex.org/I66862912"]}],"countries":["SE"],"is_corresponding":false,"raw_author_name":"Fareed Qararyah","raw_affiliation_strings":["Chalmers University of Technology, Sweden"],"raw_orcid":"https://orcid.org/0000-0002-3955-2836","affiliations":[{"raw_affiliation_string":"Chalmers University of Technology, Sweden","institution_ids":["https://openalex.org/I66862912"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008522376","display_name":"Muhammad Waqar Azhar","orcid":"https://orcid.org/0000-0003-0477-4540"},"institutions":[{"id":"https://openalex.org/I66862912","display_name":"Chalmers University of Technology","ror":"https://ror.org/040wg7k59","country_code":"SE","type":"education","lineage":["https://openalex.org/I66862912"]}],"countries":["SE"],"is_corresponding":false,"raw_author_name":"Muhammad Waqar Azhar","raw_affiliation_strings":["Chalmers University of Technology, Sweden"],"raw_orcid":"https://orcid.org/0000-0003-0477-4540","affiliations":[{"raw_affiliation_string":"Chalmers University of Technology, Sweden","institution_ids":["https://openalex.org/I66862912"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5043427541","display_name":"Pedro Trancoso","orcid":"https://orcid.org/0000-0002-2776-9253"},"institutions":[{"id":"https://openalex.org/I66862912","display_name":"Chalmers University of Technology","ror":"https://ror.org/040wg7k59","country_code":"SE","type":"education","lineage":["https://openalex.org/I66862912"]}],"countries":["SE"],"is_corresponding":false,"raw_author_name":"Pedro Trancoso","raw_affiliation_strings":["Chalmers University of Technology, Sweden"],"raw_orcid":"https://orcid.org/0000-0002-2776-9253","affiliations":[{"raw_affiliation_string":"Chalmers University of Technology, Sweden","institution_ids":["https://openalex.org/I66862912"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I66862912"],"apc_list":null,"apc_paid":null,"fwci":1.1564,"has_fulltext":true,"cited_by_count":7,"citation_normalized_percentile":{"value":0.75879802,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":98},"biblio":{"volume":"21","issue":"2","first_page":"1","last_page":"26"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9993000030517578,"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":0.9993000030517578,"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/T12357","display_name":"Digital Media Forensic Detection","score":0.9944000244140625,"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/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.9941999912261963,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.787940263748169},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.6070254445075989},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4910423755645752},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.32513874769210815}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.787940263748169},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.6070254445075989},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4910423755645752},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32513874769210815}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1145/3639823","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3639823","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3639823","source":{"id":"https://openalex.org/S26056741","display_name":"ACM Transactions on Architecture and Code Optimization","issn_l":"1544-3566","issn":["1544-3566","1544-3973"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Architecture and Code Optimization","raw_type":"journal-article"},{"id":"pmh:oai:research.chalmers.se:541391","is_oa":true,"landing_page_url":"https://research.chalmers.se/en/publication/a7ee32e6-0474-4d5f-9eaa-8fdd9671e56f","pdf_url":"https://research.chalmers.se/publication/541391/file/541391_Fulltext.pdf","source":{"id":"https://openalex.org/S4306402469","display_name":"Chalmers Research (Chalmers University of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I66862912","host_organization_name":"Chalmers University of Technology","host_organization_lineage":["https://openalex.org/I66862912"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":null},{"id":"pmh:oai:research.chalmers.se:539325","is_oa":false,"landing_page_url":"https://research.chalmers.se/en/publication/539325","pdf_url":null,"source":{"id":"https://openalex.org/S4306402469","display_name":"Chalmers Research (Chalmers University of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I66862912","host_organization_name":"Chalmers University of Technology","host_organization_lineage":["https://openalex.org/I66862912"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":null}],"best_oa_location":{"id":"doi:10.1145/3639823","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3639823","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3639823","source":{"id":"https://openalex.org/S26056741","display_name":"ACM Transactions on Architecture and Code Optimization","issn_l":"1544-3566","issn":["1544-3566","1544-3973"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Architecture and Code Optimization","raw_type":"journal-article"},"sustainable_development_goals":[{"score":0.8199999928474426,"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy"}],"awards":[{"id":"https://openalex.org/G4255628366","display_name":null,"funder_award_id":"957197","funder_id":"https://openalex.org/F4320320940","funder_display_name":"Stiftelsen f\u00f6r\u00a0Strategisk Forskning"},{"id":"https://openalex.org/G5185518776","display_name":"Very Efficient Deep Learning in IOT","funder_award_id":"957197","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"},{"id":"https://openalex.org/G979276111","display_name":"PRIDE: Principer f\u00f6r ber\u00e4knande minnesenheter","funder_award_id":"CHI19-0048","funder_id":"https://openalex.org/F4320320940","funder_display_name":"Stiftelsen f\u00f6r\u00a0Strategisk Forskning"}],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320320940","display_name":"Stiftelsen f\u00f6r\u00a0Strategisk Forskning","ror":"https://ror.org/044wr7g58"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4390658705.pdf","grobid_xml":"https://content.openalex.org/works/W4390658705.grobid-xml"},"referenced_works_count":61,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W2015861736","https://openalex.org/W2091158003","https://openalex.org/W2097117768","https://openalex.org/W2108598243","https://openalex.org/W2163605009","https://openalex.org/W2194775991","https://openalex.org/W2276486856","https://openalex.org/W2279098554","https://openalex.org/W2442974303","https://openalex.org/W2513568085","https://openalex.org/W2531409750","https://openalex.org/W2565125333","https://openalex.org/W2584311934","https://openalex.org/W2604319603","https://openalex.org/W2605347906","https://openalex.org/W2606722458","https://openalex.org/W2612445135","https://openalex.org/W2618530766","https://openalex.org/W2625954420","https://openalex.org/W2626991402","https://openalex.org/W2731888263","https://openalex.org/W2741581557","https://openalex.org/W2774029808","https://openalex.org/W2887936511","https://openalex.org/W2891946740","https://openalex.org/W2902251695","https://openalex.org/W2919115771","https://openalex.org/W2919512338","https://openalex.org/W2963122961","https://openalex.org/W2963125010","https://openalex.org/W2963163009","https://openalex.org/W2963918968","https://openalex.org/W2964137095","https://openalex.org/W2964259004","https://openalex.org/W2989331028","https://openalex.org/W2990714382","https://openalex.org/W3004242689","https://openalex.org/W3006336861","https://openalex.org/W3064585691","https://openalex.org/W3105234421","https://openalex.org/W3152980481","https://openalex.org/W3158057740","https://openalex.org/W3168376295","https://openalex.org/W3173944528","https://openalex.org/W3206210071","https://openalex.org/W4214604703","https://openalex.org/W4232418947","https://openalex.org/W4247951052","https://openalex.org/W4249459551","https://openalex.org/W4251309856","https://openalex.org/W4253807463","https://openalex.org/W4292794768","https://openalex.org/W4297775537","https://openalex.org/W4300337599","https://openalex.org/W4301485103","https://openalex.org/W4313194044","https://openalex.org/W4313291090","https://openalex.org/W4362644923","https://openalex.org/W6713132643","https://openalex.org/W6968655253"],"related_works":["https://openalex.org/W2731899572","https://openalex.org/W2961085424","https://openalex.org/W3215138031","https://openalex.org/W4306674287","https://openalex.org/W3009238340","https://openalex.org/W4321369474","https://openalex.org/W4360585206","https://openalex.org/W4285208911","https://openalex.org/W3046775127","https://openalex.org/W3082895349"],"abstract_inverted_index":{"Resource-efficient":[0],"Convolutional":[1],"Neural":[2],"Networks":[3],"(CNNs)":[4],"are":[5,107],"gaining":[6],"more":[7,26,181,187],"attention.":[8],"These":[9],"CNNs":[10,23],"have":[11,58],"relatively":[12],"low":[13],"computational":[14],"and":[15,40,109,168,189,206,217,246,253,261,283],"memory":[16,260,278],"requirements.":[17],"A":[18],"common":[19],"denominator":[20],"among":[21],"such":[22,105,122],"is":[24,33,74,113,215],"having":[25],"heterogeneity":[27,32,182],"than":[28,183,191],"traditional":[29,292],"CNNs.":[30],"This":[31,94],"present":[34],"at":[35],"two":[36],"levels:":[37],"intra-layer":[38],"type":[39],"inter-layer":[41],"type.":[42,131],"Generic":[43],"accelerators":[44,61,85,106,136,142],"do":[45],"not":[46,147],"capture":[47,143],"these":[48,135],"levels":[49],"of":[50,119,129,164,176],"heterogeneity,":[51],"which":[52],"harms":[53],"their":[54],"efficiency.":[55],"Consequently,":[56],"researchers":[57],"proposed":[59,213],"model-specific":[60],"with":[62,69,87],"dedicated":[63,70,120],"engines.":[64],"When":[65],"designing":[66],"an":[67,165,169,266],"accelerator":[68,162],"engines,":[71],"one":[72,77,130],"option":[73,112],"to":[75,84,114,134,243,250,265,281,288,291],"dedicate":[76],"engine":[78,99,125],"per":[79],"CNN":[80,157,177],"layer.":[81,103],"We":[82,132,151],"refer":[83,133],"designed":[86],"this":[88],"approach":[89,95],"as":[90,137],"single-engine":[91,138],"single-layer":[92],"(SESL).":[93],"enables":[96],"optimizing":[97],"each":[98,124,172],"for":[100],"its":[101],"specific":[102],"However,":[104],"resource-demanding":[108],"unscalable.":[110],"Another":[111],"design":[115],"a":[116,160,174,196,204],"minimal":[117],"number":[118],"engines":[121],"that":[123,236,273],"handles":[126],"all":[127],"layers":[128],"multiple-layer":[139],"(SEML).":[140],"SEML":[141,170,184,254,267],"the":[144,148,153,239,276],"inter-layer-type":[145],"but":[146],"intra-layer-type":[149],"heterogeneity.":[150],"propose":[152,195],"Fixed":[154],"Budget":[155],"Hybrid":[156],"Accelerator":[158],"(FiBHA),":[159],"hybrid":[161],"composed":[163],"SESL":[166,208,252],"part":[167],"part,":[171],"processing":[173],"subset":[175],"layers.":[178],"FiBHA":[179,237,258],"captures":[180],"while":[185],"being":[186],"resource-aware":[188],"scalable":[190],"SESL.":[192],"Moreover,":[193,257],"we":[194],"novel":[197],"module,":[198],"Fused":[199],"Inverted":[200],"Residual":[201],"Bottleneck":[202],"(FIRB),":[203],"fine-grained":[205],"memory-light":[207],"architecture":[209,214],"building":[210],"block.":[211],"The":[212,269],"implemented":[216],"evaluated":[218],"using":[219],"high-level":[220],"synthesis":[221],"(HLS)":[222],"on":[223],"different":[224],"Field":[225],"Programmable":[226],"Gate":[227],"Arrays":[228],"representing":[229],"various":[230],"resource":[231],"budgets.":[232],"Our":[233],"evaluation":[234,270],"shows":[235,272],"improves":[238],"throughput":[240],"by":[241,279,286],"up":[242,280,287],"4":[244],"x":[245,248],"2.5":[247],"compared":[249,264,290],"state-of-the-art":[251],"accelerators,":[255],"respectively.":[256],"reduces":[259,275],"energy":[262,284],"consumption":[263],"accelerator.":[268],"also":[271],"FIRB":[274],"required":[277],"54%,":[282],"requirements":[285],"35%":[289],"pipelining.":[293]},"counts_by_year":[{"year":2025,"cited_by_count":5},{"year":2024,"cited_by_count":2}],"updated_date":"2026-07-20T07:56:41.581041","created_date":"2024-01-13T00:00:00"}