{"id":"https://openalex.org/W4411725710","doi":"https://doi.org/10.1109/iscas56072.2025.11043768","title":"Sparsity-Aware Optimization of In-Memory Bayesian Binary Neural Network Accelerators","display_name":"Sparsity-Aware Optimization of In-Memory Bayesian Binary Neural Network Accelerators","publication_year":2025,"publication_date":"2025-05-25","ids":{"openalex":"https://openalex.org/W4411725710","doi":"https://doi.org/10.1109/iscas56072.2025.11043768"},"language":"en","primary_location":{"id":"doi:10.1109/iscas56072.2025.11043768","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas56072.2025.11043768","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on Circuits and Systems (ISCAS)","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/A5076449772","display_name":"Prabodh Katti","orcid":"https://orcid.org/0000-0003-1919-9593"},"institutions":[{"id":"https://openalex.org/I183935753","display_name":"King's College London","ror":"https://ror.org/0220mzb33","country_code":"GB","type":"education","lineage":["https://openalex.org/I124357947","https://openalex.org/I183935753"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Prabodh Katti","raw_affiliation_strings":["King’s College London,Department of Engineering,London,UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"King’s College London,Department of Engineering,London,UK","institution_ids":["https://openalex.org/I183935753"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012783672","display_name":"Bashir M. Al\u2010Hashimi","orcid":"https://orcid.org/0000-0002-3591-1328"},"institutions":[{"id":"https://openalex.org/I183935753","display_name":"King's College London","ror":"https://ror.org/0220mzb33","country_code":"GB","type":"education","lineage":["https://openalex.org/I124357947","https://openalex.org/I183935753"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Bashir M. Al-Hashimi","raw_affiliation_strings":["King’s College London,Department of Engineering,London,UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"King’s College London,Department of Engineering,London,UK","institution_ids":["https://openalex.org/I183935753"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5048595299","display_name":"Bipin Rajendran","orcid":"https://orcid.org/0000-0002-2960-6909"},"institutions":[{"id":"https://openalex.org/I183935753","display_name":"King's College London","ror":"https://ror.org/0220mzb33","country_code":"GB","type":"education","lineage":["https://openalex.org/I124357947","https://openalex.org/I183935753"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Bipin Rajendran","raw_affiliation_strings":["King’s College London,Department of Engineering,London,UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"King’s College London,Department of Engineering,London,UK","institution_ids":["https://openalex.org/I183935753"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.5175,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.92807456,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.9940000176429749,"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"}},"topics":[{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.9940000176429749,"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"}},{"id":"https://openalex.org/T10876","display_name":"Fault Detection and Control Systems","score":0.9846000075340271,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","score":0.965399980545044,"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":0.7264466285705566},{"id":"https://openalex.org/keywords/bayesian-optimization","display_name":"Bayesian optimization","score":0.5830762982368469},{"id":"https://openalex.org/keywords/binary-number","display_name":"Binary number","score":0.5572070479393005},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5351352095603943},{"id":"https://openalex.org/keywords/bayesian-network","display_name":"Bayesian network","score":0.4923752546310425},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.46845126152038574},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3759498596191406},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.357791006565094},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10196369886398315},{"id":"https://openalex.org/keywords/arithmetic","display_name":"Arithmetic","score":0.07808583974838257}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7264466285705566},{"id":"https://openalex.org/C2778049539","wikidata":"https://www.wikidata.org/wiki/Q17002908","display_name":"Bayesian optimization","level":2,"score":0.5830762982368469},{"id":"https://openalex.org/C48372109","wikidata":"https://www.wikidata.org/wiki/Q3913","display_name":"Binary number","level":2,"score":0.5572070479393005},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5351352095603943},{"id":"https://openalex.org/C33724603","wikidata":"https://www.wikidata.org/wiki/Q812540","display_name":"Bayesian network","level":2,"score":0.4923752546310425},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.46845126152038574},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3759498596191406},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.357791006565094},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10196369886398315},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.07808583974838257}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iscas56072.2025.11043768","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas56072.2025.11043768","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W2897686304","https://openalex.org/W2948661249","https://openalex.org/W2953352640","https://openalex.org/W2964059111","https://openalex.org/W2987612038","https://openalex.org/W3005619596","https://openalex.org/W3043426275","https://openalex.org/W3090581962","https://openalex.org/W3152980481","https://openalex.org/W3197327073","https://openalex.org/W3211621070","https://openalex.org/W4251999575","https://openalex.org/W4285144887","https://openalex.org/W4292169167","https://openalex.org/W4317927903","https://openalex.org/W4319068960","https://openalex.org/W4321001968","https://openalex.org/W4385245195","https://openalex.org/W4388822849","https://openalex.org/W4400231467","https://openalex.org/W4401548024","https://openalex.org/W6664870454","https://openalex.org/W6684488266","https://openalex.org/W6772650050","https://openalex.org/W6773104461","https://openalex.org/W6773599722","https://openalex.org/W6779746701","https://openalex.org/W6787972765","https://openalex.org/W6846590345"],"related_works":["https://openalex.org/W3106461837","https://openalex.org/W4283700523","https://openalex.org/W3168182983","https://openalex.org/W4387079005","https://openalex.org/W4321472004","https://openalex.org/W4221166418","https://openalex.org/W2968265130","https://openalex.org/W2973085194","https://openalex.org/W3117893869","https://openalex.org/W4401858220"],"abstract_inverted_index":{"Bayesian":[0,90],"Neural":[1,92],"Networks":[2],"(BNNs)":[3],"provide":[4],"principled":[5],"estimates":[6],"of":[7,76,105,111,118,146,152,165],"model":[8],"and":[9,50,78,123,230,239,252],"data":[10],"uncertainty":[11,189],"by":[12,41,178],"encoding":[13],"parameters":[14,160,166],"as":[15,156,158],"distributions.":[16],"This":[17],"makes":[18],"them":[19],"key":[20],"enablers":[21],"for":[22,69,89,192,232],"reliable":[23],"AI":[24],"that":[25,96,113,136],"can":[26,36],"be":[27,37],"deployed":[28],"on":[29,196,200],"safety":[30],"critical":[31],"edge":[32],"systems.":[33],"These":[34],"systems":[35],"made":[38,109],"resource":[39],"efficient":[40],"restricting":[42],"synapses":[43,112],"to":[44,180,219,222,258,272],"two":[45],"synaptic":[46],"states":[47],"{\u22121,":[48],"+1},":[49],"using":[51],"a":[52,85,115,143,150,162,193,201,214,259],"memristive":[53],"in-memory":[54],"computing":[55],"(IMC)":[56],"paradigm.":[57],"However,":[58],"BNNs":[59],"pose":[60],"an":[61],"additional":[62],"challenge":[63],"\u2013":[64,103],"they":[65],"require":[66,124,170],"multiple":[67],"instantiations":[68],"ensembling,":[70],"consuming":[71],"extra":[72],"resources":[73],"in":[74,167,186,236,250,255,266],"terms":[75],"energy":[77,256],"area.":[79],"In":[80],"this":[81],"work,":[82],"we":[83,226,245],"propose":[84],"novel":[86],"sparsity-aware":[87,203,233,243],"optimization":[88,128],"Binary":[91],"Network":[93],"(BBNN)":[94],"accelerators":[95],"exploits":[97,132],"the":[98,106,133,147,153,159,273],"inherent":[99],"BBNN":[100],"sampling":[101,134],"sparsity":[102,135],"most":[104],"network":[107,148,195],"is":[108],"up":[110,179],"have":[114],"high":[116],"probability":[117],"being":[119],"fixed":[120],"at":[121],"\u00b11":[122],"no":[125,184],"sampling.":[126],"The":[127],"scheme":[129],"proposed":[130],"here":[131],"exists":[137],"both":[138,237],"among":[139],"layers,":[140],"i.e.,":[141,161],"only":[142],"few":[144],"layers":[145,169],"contain":[149],"majority":[151],"probabilistic":[154],"synapses,":[155],"well":[157],"tiny":[163],"fraction":[164],"these":[168],"sampling,":[171],"reducing":[172],"total":[173],"sampled":[174],"parameter":[175],"count":[176],"further":[177],"86%.":[181],"We":[182,211],"demonstrate":[183,220],"loss":[185],"accuracy":[187],"or":[188],"quantification":[190],"performance":[191],"VGGBinaryConnect":[194],"CIFAR-100":[197],"dataset":[198],"mapped":[199],"custom":[202],"phase":[204],"change":[205],"memory":[206],"(PCM)":[207],"based":[208],"IMC":[209],"simulator.":[210],"also":[212,264],"develop":[213],"simple":[215],"drift":[216],"compensation":[217],"technique":[218],"robustness":[221],"drift-induced":[223],"degradation.":[224],"Finally,":[225],"project":[227],"latency,":[228],"energy,":[229],"area":[231,251],"BNN":[234,275],"implementation":[235],"pipelined":[238],"non-pipelined":[240],"modes.":[241],"With":[242],"implementation,":[244],"estimate":[246],"upto":[247],"5.3\u00d7":[248],"reduction":[249,254],"8.8\u00d7":[253],"compared":[257,271],"non-sparsity-aware":[260],"implementation.":[261],"Our":[262],"approach":[263],"results":[265],"2.9\u00d7":[267],"more":[268],"power":[269],"efficiency":[270],"state-of-the-art":[274],"accelerator.":[276]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}