{"id":"https://openalex.org/W7130614757","doi":"https://doi.org/10.1109/ictc66702.2025.11389078","title":"Accelerating CNN Inference on MCUs with Quantized Early-Exit Networks","display_name":"Accelerating CNN Inference on MCUs with Quantized Early-Exit Networks","publication_year":2025,"publication_date":"2025-10-14","ids":{"openalex":"https://openalex.org/W7130614757","doi":"https://doi.org/10.1109/ictc66702.2025.11389078"},"language":null,"primary_location":{"id":"doi:10.1109/ictc66702.2025.11389078","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ictc66702.2025.11389078","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 16th International Conference on Information and Communication Technology Convergence (ICTC)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5076301234","display_name":"Gunju Park","orcid":"https://orcid.org/0000-0002-6734-8648"},"institutions":[{"id":"https://openalex.org/I142401562","display_name":"Electronics and Telecommunications Research Institute","ror":"https://ror.org/03ysstz10","country_code":"KR","type":"facility","lineage":["https://openalex.org/I142401562","https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Gunju Park","raw_affiliation_strings":["Electronics and Telecommunications Research Institute,Daejeon,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Electronics and Telecommunications Research Institute,Daejeon,Republic of Korea","institution_ids":["https://openalex.org/I142401562"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054768260","display_name":"Seungtae Hong","orcid":"https://orcid.org/0000-0003-0273-3542"},"institutions":[{"id":"https://openalex.org/I142401562","display_name":"Electronics and Telecommunications Research Institute","ror":"https://ror.org/03ysstz10","country_code":"KR","type":"facility","lineage":["https://openalex.org/I142401562","https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seungtae Hong","raw_affiliation_strings":["Electronics and Telecommunications Research Institute,Daejeon,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Electronics and Telecommunications Research Institute,Daejeon,Republic of Korea","institution_ids":["https://openalex.org/I142401562"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5062319303","display_name":"Jeong-Si Kim","orcid":null},"institutions":[{"id":"https://openalex.org/I142401562","display_name":"Electronics and Telecommunications Research Institute","ror":"https://ror.org/03ysstz10","country_code":"KR","type":"facility","lineage":["https://openalex.org/I142401562","https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jeong-Si Kim","raw_affiliation_strings":["Electronics and Telecommunications Research Institute,Daejeon,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Electronics and Telecommunications Research Institute,Daejeon,Republic of Korea","institution_ids":["https://openalex.org/I142401562"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I142401562"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"389","last_page":"391"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.5152999758720398,"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":0.5152999758720398,"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/T10363","display_name":"Low-power high-performance VLSI design","score":0.06419999897480011,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.047200001776218414,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.7402999997138977},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.7009999752044678},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6924999952316284},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.6047000288963318},{"id":"https://openalex.org/keywords/microcontroller","display_name":"Microcontroller","score":0.5814999938011169},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4652000069618225},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.4431000053882599},{"id":"https://openalex.org/keywords/static-random-access-memory","display_name":"Static random-access memory","score":0.4253999888896942}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.7402999997138977},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7371000051498413},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.7009999752044678},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6924999952316284},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.6047000288963318},{"id":"https://openalex.org/C173018170","wikidata":"https://www.wikidata.org/wiki/Q165678","display_name":"Microcontroller","level":2,"score":0.5814999938011169},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4652000069618225},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.4431000053882599},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4309999942779541},{"id":"https://openalex.org/C68043766","wikidata":"https://www.wikidata.org/wiki/Q267416","display_name":"Static random-access memory","level":2,"score":0.4253999888896942},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3919999897480011},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.35089999437332153},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.3481000065803528},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.3156999945640564},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3156000077724457},{"id":"https://openalex.org/C46743427","wikidata":"https://www.wikidata.org/wiki/Q1341685","display_name":"Inference engine","level":3,"score":0.3100000023841858},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.3095000088214874},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.29829999804496765},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.2867000102996826},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.2856000065803528},{"id":"https://openalex.org/C157170001","wikidata":"https://www.wikidata.org/wiki/Q4781507","display_name":"Applications of artificial intelligence","level":2,"score":0.2793999910354614},{"id":"https://openalex.org/C164155591","wikidata":"https://www.wikidata.org/wiki/Q2067766","display_name":"Satisfiability modulo theories","level":2,"score":0.2635999917984009}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ictc66702.2025.11389078","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ictc66702.2025.11389078","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 16th International Conference on Information and Communication Technology Convergence (ICTC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320334874","display_name":"Defense Acquisition Program Administration","ror":"https://ror.org/04bjg9m96"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":3,"referenced_works":["https://openalex.org/W2963122961","https://openalex.org/W2963163009","https://openalex.org/W4290043197"],"related_works":[],"abstract_inverted_index":{"On-device":[0],"AI":[1],"on":[2,60,119],"microcontrollers":[3],"(MCUs)":[4],"is":[5],"severely":[6],"hampered":[7],"by":[8,49],"their":[9],"limited":[10],"memory":[11],"and":[12,25,74,79,113],"computational":[13],"resources.":[14],"To":[15],"address":[16],"this,":[17],"we":[18,81],"propose":[19],"a":[20,34,44,61,65,72,85,91,96,111],"complete":[21],"pipeline":[22],"for":[23,116],"creating":[24],"deploying":[26],"efficient,":[27],"dynamic":[28],"neural":[29],"networks.":[30],"Our":[31],"methodology":[32],"features":[33],"mixed":[35],"training":[36],"strategy":[37],"to":[38,43,84,90],"stably":[39],"attach":[40],"early-exit":[41,108],"branches":[42],"pretrained":[45],"CNN":[46],"backbone,":[47],"followed":[48],"INT8":[50],"post-training":[51],"quantization":[52],"(PTQ).":[53],"We":[54],"validated":[55],"our":[56],"approach":[57],"with":[58,95],"TensorFlow-Lite":[59],"host":[62],"CPU":[63],"using":[64],"Modified":[66],"MobileNetV2":[67],"architecture.":[68],"The":[69],"results":[70],"demonstrate":[71],"clear":[73],"controllable":[75],"trade-off":[76],"between":[77],"latency":[78],"accuracy;":[80],"achieve":[82],"up":[83],"1.54x":[86],"inference":[87,118],"speedup":[88],"compared":[89],"standard":[92],"quantized":[93,107],"model,":[94],"peak":[97],"SRAM":[98],"usage":[99],"of":[100],"only":[101],"67KB.":[102],"This":[103],"work":[104],"confirms":[105],"that":[106],"networks":[109],"are":[110],"practical":[112],"effective":[114],"solution":[115],"accelerating":[117],"resource-constrained":[120],"devices.":[121]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-02-20T00:00:00"}