{"id":"https://openalex.org/W4405935318","doi":"https://doi.org/10.1109/tcasai.2024.3523858","title":"On-Chip Memory Optimization for Deep Learning Inference on Resource-Deficient Devices","display_name":"On-Chip Memory Optimization for Deep Learning Inference on Resource-Deficient Devices","publication_year":2024,"publication_date":"2024-12-31","ids":{"openalex":"https://openalex.org/W4405935318","doi":"https://doi.org/10.1109/tcasai.2024.3523858"},"language":"en","primary_location":{"id":"doi:10.1109/tcasai.2024.3523858","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcasai.2024.3523858","pdf_url":null,"source":{"id":"https://openalex.org/S4404675360","display_name":"IEEE transactions on circuits and systems for artificial intelligence.","issn_l":"2996-6647","issn":["2996-6647"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Circuits and Systems for Artificial Intelligence","raw_type":"journal-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":null,"display_name":"Deepak Kumar","orcid":"https://orcid.org/0009-0008-2840-0133"},"institutions":[{"id":"https://openalex.org/I132153292","display_name":"Indian Institute of Technology Patna","ror":"https://ror.org/01ft5vz71","country_code":"IN","type":"education","lineage":["https://openalex.org/I132153292"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Deepak Kumar","raw_affiliation_strings":["Department of Electrical Engineering, Indian Institute of Technology Patna, Patna, India","Department of Electrical Engineering, Indian Institute of Technology Patna, Bihta, Patna, India"],"raw_orcid":"https://orcid.org/0009-0008-2840-0133","affiliations":[{"raw_affiliation_string":"Department of Electrical Engineering, Indian Institute of Technology Patna, Patna, India","institution_ids":["https://openalex.org/I132153292"]},{"raw_affiliation_string":"Department of Electrical Engineering, Indian Institute of Technology Patna, Bihta, Patna, India","institution_ids":["https://openalex.org/I132153292"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5056883826","display_name":"Udit Satija","orcid":"https://orcid.org/0000-0002-3851-765X"},"institutions":[{"id":"https://openalex.org/I132153292","display_name":"Indian Institute of Technology Patna","ror":"https://ror.org/01ft5vz71","country_code":"IN","type":"education","lineage":["https://openalex.org/I132153292"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Udit Satija","raw_affiliation_strings":["Department of Electrical Engineering, Indian Institute of Technology Patna, Patna, India","Department of Electrical Engineering, Indian Institute of Technology Patna, Bihta, Patna, India"],"raw_orcid":"https://orcid.org/0000-0002-3851-765X","affiliations":[{"raw_affiliation_string":"Department of Electrical Engineering, Indian Institute of Technology Patna, Patna, India","institution_ids":["https://openalex.org/I132153292"]},{"raw_affiliation_string":"Department of Electrical Engineering, Indian Institute of Technology Patna, Bihta, Patna, India","institution_ids":["https://openalex.org/I132153292"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I132153292"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.25102663,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"2","issue":"1","first_page":"14","last_page":"24"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9768000245094299,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9768000245094299,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9284999966621399,"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"}},{"id":"https://openalex.org/T10363","display_name":"Low-power high-performance VLSI design","score":0.919700026512146,"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/inference","display_name":"Inference","score":0.7493271827697754},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.5881426930427551},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5678961277008057},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.523069441318512},{"id":"https://openalex.org/keywords/chip","display_name":"Chip","score":0.5013875961303711},{"id":"https://openalex.org/keywords/resource","display_name":"Resource (disambiguation)","score":0.4903900623321533},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.41142308712005615},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.3686068058013916},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.09390866756439209},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.08312815427780151}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7493271827697754},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.5881426930427551},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5678961277008057},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.523069441318512},{"id":"https://openalex.org/C165005293","wikidata":"https://www.wikidata.org/wiki/Q1074500","display_name":"Chip","level":2,"score":0.5013875961303711},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.4903900623321533},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.41142308712005615},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3686068058013916},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.09390866756439209},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.08312815427780151}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tcasai.2024.3523858","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcasai.2024.3523858","pdf_url":null,"source":{"id":"https://openalex.org/S4404675360","display_name":"IEEE transactions on circuits and systems for artificial intelligence.","issn_l":"2996-6647","issn":["2996-6647"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Circuits and Systems for Artificial Intelligence","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":36,"referenced_works":["https://openalex.org/W1485846902","https://openalex.org/W1859757340","https://openalex.org/W2300242332","https://openalex.org/W2308736390","https://openalex.org/W2928560789","https://openalex.org/W2950656546","https://openalex.org/W2999093589","https://openalex.org/W3007729836","https://openalex.org/W3021575660","https://openalex.org/W3137762252","https://openalex.org/W3138439374","https://openalex.org/W3154584980","https://openalex.org/W3166384310","https://openalex.org/W3175548485","https://openalex.org/W3192682950","https://openalex.org/W4226319959","https://openalex.org/W4295783852","https://openalex.org/W4306250145","https://openalex.org/W4309287541","https://openalex.org/W4312817580","https://openalex.org/W4366351848","https://openalex.org/W4385892121","https://openalex.org/W4392207810","https://openalex.org/W4392903807","https://openalex.org/W6617368339","https://openalex.org/W6637709462","https://openalex.org/W6663865803","https://openalex.org/W6677580257","https://openalex.org/W6683826617","https://openalex.org/W6725543821","https://openalex.org/W6726275242","https://openalex.org/W6729956949","https://openalex.org/W6744700018","https://openalex.org/W6754873534","https://openalex.org/W6762935145","https://openalex.org/W6780827055"],"related_works":["https://openalex.org/W4375867731","https://openalex.org/W2055243143","https://openalex.org/W2611989081","https://openalex.org/W2731899572","https://openalex.org/W4230611425","https://openalex.org/W4294635752","https://openalex.org/W4304166257","https://openalex.org/W4383066092","https://openalex.org/W3215138031","https://openalex.org/W1986418932"],"abstract_inverted_index":{"Advances":[0],"in":[1,108],"deep":[2,59],"learning":[3],"(DL)":[4],"have":[5],"enabled":[6],"the":[7,44,58,74,100,115,124,127,135,142,163],"integration":[8],"of":[9,126,149,154,162],"intelligence":[10],"into":[11,65],"low-end":[12],"Internet-of-Things":[13],"(IoT)":[14],"devices.":[15],"However,":[16],"traditional":[17],"DL":[18,45],"inference":[19],"on":[20,47,130],"resource-constrained":[21],"microcontrollers":[22],"(MCUs)":[23],"faces":[24],"significant":[25],"memory":[26,41,72,91,171],"and":[27,36,93,151,160,177,181],"computational":[28],"challenges.":[29],"This":[30],"article":[31],"presents":[32],"a":[33,52,80,131],"novel":[34,81],"compression":[35],"reconstruction":[37,84],"method":[38,129,140],"for":[39,87,188],"on-chip":[40],"optimization":[42],"enabling":[43],"model":[46,143],"such":[48],"MCUs.":[49],"We":[50,122],"exploit":[51],"low-rank":[53],"approximation":[54],"technique":[55,86],"to":[56,147],"decompose":[57],"neural":[60,178],"network":[61],"(DNN)":[62],"weight":[63,118],"matrix":[64,119],"lower-rank":[66],"sub-matrices":[67],"that":[68],"occupy":[69],"less":[70,170],"Flash":[71,150],"than":[73,172],"uncompressed":[75],"matrix.":[76],"Further,":[77],"we":[78],"propose":[79],"layer-wise":[82],"kernel":[83,102],"(LaWKeR)":[85],"static":[88],"random":[89],"access":[90],"(SRAM)":[92],"FLOPS":[94],"reduction.":[95],"The":[96,138],"LaWKeR":[97,139],"reconstructs":[98],"only":[99,155,158],"active":[101],"directly":[103],"from":[104,145],"compressed":[105,117],"weights":[106],"stored":[107],"Flash,":[109],"unlike":[110],"existing":[111],"methods":[112],"which":[113],"reconstruct":[114],"whole":[116],"during":[120],"run-time.":[121],"demonstrate":[123],"efficacy":[125],"proposed":[128],"NUCLEO":[132],"board":[133],"using":[134,157],"STM32F401":[136],"microcontroller.":[137],"reduces":[141],"size":[144],"73KB":[146],"20.4KB":[148],"RAM":[152],"usage":[153],"21.34KB,":[156],"4.20%":[159],"21.28%":[161],"total":[164],"embedded":[165],"memory,":[166],"respectively.":[167],"It":[168],"requires":[169],"popular":[173],"frameworks":[174],"like":[175],"tinyML":[176],"architecture":[179],"search":[180],"can":[182],"be":[183],"combined":[184],"with":[185],"other":[186],"techniques":[187],"even":[189],"smaller":[190],"models.":[191]},"counts_by_year":[],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}