{"id":"https://openalex.org/W4408888778","doi":"https://doi.org/10.1145/3689031.3696067","title":"Flex: Fast, Accurate DNN Inference on Low-Cost Edges Using Heterogeneous Accelerator Execution","display_name":"Flex: Fast, Accurate DNN Inference on Low-Cost Edges Using Heterogeneous Accelerator Execution","publication_year":2025,"publication_date":"2025-03-26","ids":{"openalex":"https://openalex.org/W4408888778","doi":"https://doi.org/10.1145/3689031.3696067"},"language":"en","primary_location":{"id":"doi:10.1145/3689031.3696067","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3689031.3696067","pdf_url":null,"source":null,"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Twentieth European Conference on Computer Systems","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3689031.3696067","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5071074000","display_name":"Tanmoy Sen","orcid":"https://orcid.org/0000-0001-7677-3358"},"institutions":[{"id":"https://openalex.org/I51556381","display_name":"University of Virginia","ror":"https://ror.org/0153tk833","country_code":"US","type":"education","lineage":["https://openalex.org/I51556381"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Tanmoy Sen","raw_affiliation_strings":["University of Virginia"],"affiliations":[{"raw_affiliation_string":"University of Virginia","institution_ids":["https://openalex.org/I51556381"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050569064","display_name":"Haiying Shen","orcid":"https://orcid.org/0000-0002-7548-6223"},"institutions":[{"id":"https://openalex.org/I51556381","display_name":"University of Virginia","ror":"https://ror.org/0153tk833","country_code":"US","type":"education","lineage":["https://openalex.org/I51556381"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Haiying Shen","raw_affiliation_strings":["University of Virginia"],"affiliations":[{"raw_affiliation_string":"University of Virginia","institution_ids":["https://openalex.org/I51556381"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5111976348","display_name":"A. Iyer","orcid":null},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Anand Padmanabha Iyer","raw_affiliation_strings":["Georgia Institute of Technology"],"affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology","institution_ids":["https://openalex.org/I130701444"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5071074000"],"corresponding_institution_ids":["https://openalex.org/I51556381"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.04408852,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"507","last_page":"523"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9965000152587891,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9950000047683716,"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/flex","display_name":"FLEX","score":0.8605114221572876},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7668700814247131},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6541919708251953},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.33091408014297485},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.28401193022727966},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.06338170170783997}],"concepts":[{"id":"https://openalex.org/C2776252893","wikidata":"https://www.wikidata.org/wiki/Q1364836","display_name":"FLEX","level":2,"score":0.8605114221572876},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7668700814247131},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6541919708251953},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.33091408014297485},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.28401193022727966},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.06338170170783997}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3689031.3696067","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3689031.3696067","pdf_url":null,"source":null,"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Twentieth European Conference on Computer Systems","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3689031.3696067","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3689031.3696067","pdf_url":null,"source":null,"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Twentieth European Conference on Computer Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2750510799","display_name":null,"funder_award_id":"2206522, 2319988","funder_id":"https://openalex.org/F4320323817","funder_display_name":"Universitas Brawijaya"}],"funders":[{"id":"https://openalex.org/F4320323817","display_name":"Universitas Brawijaya","ror":"https://ror.org/01wk3d929"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":54,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W134960717","https://openalex.org/W2013281957","https://openalex.org/W2057907879","https://openalex.org/W2108598243","https://openalex.org/W2296073425","https://openalex.org/W2613019972","https://openalex.org/W2734941459","https://openalex.org/W2781726626","https://openalex.org/W2860338957","https://openalex.org/W2885579974","https://openalex.org/W2887117815","https://openalex.org/W2911964244","https://openalex.org/W2914062858","https://openalex.org/W2931092525","https://openalex.org/W2951913055","https://openalex.org/W2956461999","https://openalex.org/W2963748441","https://openalex.org/W2969374972","https://openalex.org/W2984745954","https://openalex.org/W2997591727","https://openalex.org/W3028307607","https://openalex.org/W3100208984","https://openalex.org/W3104494818","https://openalex.org/W3106762601","https://openalex.org/W3107788976","https://openalex.org/W3135619047","https://openalex.org/W3155039760","https://openalex.org/W3157901373","https://openalex.org/W3163148797","https://openalex.org/W3164200338","https://openalex.org/W3168682287","https://openalex.org/W3172887788","https://openalex.org/W3186289964","https://openalex.org/W3196578351","https://openalex.org/W3204294171","https://openalex.org/W3204616036","https://openalex.org/W3215253865","https://openalex.org/W4200365713","https://openalex.org/W4205808347","https://openalex.org/W4213135659","https://openalex.org/W4230033334","https://openalex.org/W4232191208","https://openalex.org/W4236976245","https://openalex.org/W4244652618","https://openalex.org/W4251117921","https://openalex.org/W4285175090","https://openalex.org/W4287119399","https://openalex.org/W4289305285","https://openalex.org/W4289821779","https://openalex.org/W4294032666","https://openalex.org/W6745245109","https://openalex.org/W6798686915","https://openalex.org/W6929237974"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4250698707","https://openalex.org/W4244545186","https://openalex.org/W4247748223","https://openalex.org/W1008394927","https://openalex.org/W183173419","https://openalex.org/W4253292776","https://openalex.org/W2483231410"],"abstract_inverted_index":{"Significant":[0],"b":[1],"reakthroughs":[2],"in":[3,43],"machine":[4],"learning":[5,127],"(ML)":[6],"and":[7,24,69,75,125,136,154,165],"the":[8,67,80,134,172],"advantages":[9],"of":[10],"on-device":[11],"processing":[12],"have":[13],"led":[14],"to":[15,33,53,129,147,152,159,162],"edge":[16,39,118],"devices":[17,55],"increasingly":[18],"incorporating":[19],"accelerators":[20,28,48],"like":[21],"GPUs,":[22],"NPUs,":[23],"DSPs.":[25],"However,":[26],"these":[27],"consume":[29],"energy,":[30],"prompting":[31],"users":[32,41],"limit":[34],"their":[35],"floating-point":[36],"precision.":[37],"Many":[38],"device":[40],"are":[42],"regions":[44],"where":[45],"including":[46],"high-fidelity":[47],"is":[49,166],"too":[50],"costly,":[51],"leading":[52],"low-cost":[54],"with":[56],"low":[57,76],"precision,":[58],"sacrificing":[59],"accuracy.":[60],"Previous":[61],"work":[62],"predetermined":[63],"layer":[64,89],"assignments":[65],"between":[66],"CPU":[68,135],"accelerator":[70],"offline":[71],"for":[72,99],"high":[73],"accuracy":[74,149],"latency":[77],"without":[78],"considering":[79],"input,":[81],"but":[82],"we":[83,94],"observe":[84],"that":[85],"input":[86],"affects":[87],"optimal":[88,170],"assignment.":[90],"To":[91],"address":[92],"this,":[93],"present":[95],"Flex,":[96],"a":[97,122],"system":[98],"Fast,":[100],"Accurate":[101],"DNN":[102],"Inference":[103],"on":[104,116],"Low-Cost":[105],"Edges":[106],"using":[107],"Heterogeneous":[108],"Accelerator":[109],"eXecution.":[110],"Leveraging":[111],"common":[112],"observations":[113],"from":[114],"models":[115],"various":[117],"devices,":[119],"Flex":[120,140],"uses":[121],"lightweight":[123],"heuristic":[124],"reinforcement":[126],"(RL)":[128],"dynamically":[130],"assign":[131],"layers":[132],"across":[133],"accelerator.":[137],"Experiments":[138],"show":[139],"improves":[141],"average":[142],"inference":[143],"time":[144],"by":[145,150,157],"up":[146,151,158],"39%,":[148],"22%,":[153],"energy":[155],"consumption":[156],"61%":[160],"compared":[161],"state-of-the-art":[163],"methods,":[164],"only":[167],"4.2%":[168],"less":[169],"than":[171],"best":[173],"achievable":[174],"results.":[175]},"counts_by_year":[],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2025-10-10T00:00:00"}