{"id":"https://openalex.org/W4313483690","doi":"https://doi.org/10.48550/arxiv.2301.00330","title":"Efficient On-device Training via Gradient Filtering","display_name":"Efficient On-device Training via Gradient Filtering","publication_year":2023,"publication_date":"2023-01-01","ids":{"openalex":"https://openalex.org/W4313483690","doi":"https://doi.org/10.48550/arxiv.2301.00330"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2301.00330","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2301.00330","pdf_url":"https://arxiv.org/pdf/2301.00330","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2301.00330","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101920091","display_name":"Yuedong Yang","orcid":"https://orcid.org/0009-0004-1505-3882"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Yuedong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108578601","display_name":"Guihong Li","orcid":"https://orcid.org/0000-0001-8537-8632"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Guihong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5036227385","display_name":"Radu M\u0103rculescu","orcid":"https://orcid.org/0000-0003-1826-7646"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Marculescu, Radu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10273","display_name":"IoT and Edge/Fog Computing","score":0.9911999702453613,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T10273","display_name":"IoT and Edge/Fog Computing","score":0.9911999702453613,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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":0.9898999929428101,"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/T12702","display_name":"Brain Tumor Detection and Classification","score":0.9891999959945679,"subfield":{"id":"https://openalex.org/subfields/2808","display_name":"Neurology"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.9354712963104248},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8089649081230164},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.5475739240646362},{"id":"https://openalex.org/keywords/point","display_name":"Point (geometry)","score":0.5265669822692871},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.48065483570098877},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.4277370572090149},{"id":"https://openalex.org/keywords/segmentation","display_name":"Segmentation","score":0.4187955856323242},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4116387665271759},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3518869876861572},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.32836979627609253},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.07478612661361694}],"concepts":[{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.9354712963104248},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8089649081230164},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.5475739240646362},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.5265669822692871},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48065483570098877},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.4277370572090149},{"id":"https://openalex.org/C89600930","wikidata":"https://www.wikidata.org/wiki/Q1423946","display_name":"Segmentation","level":2,"score":0.4187955856323242},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4116387665271759},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3518869876861572},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.32836979627609253},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.07478612661361694},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2301.00330","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2301.00330","pdf_url":"https://arxiv.org/pdf/2301.00330","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2301.00330","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2301.00330","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2301.00330","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2301.00330","pdf_url":"https://arxiv.org/pdf/2301.00330","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[{"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7","score":0.8999999761581421}],"awards":[{"id":"https://openalex.org/G1154834153","display_name":"CNS Core: Small: Bringing the Cloud to the Masses: Models, Algorithms, and Systems Targeting Deep Learning for the IoT Era","funder_award_id":"2007284","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4313483690.pdf","grobid_xml":"https://content.openalex.org/works/W4313483690.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2317245370","https://openalex.org/W4249323025","https://openalex.org/W198851386","https://openalex.org/W2030310580","https://openalex.org/W947442053","https://openalex.org/W1980160788","https://openalex.org/W2148915962","https://openalex.org/W2283866686","https://openalex.org/W4287182096","https://openalex.org/W27867058"],"abstract_inverted_index":{"Despite":[0],"its":[1],"importance":[2],"for":[3,18,194],"federated":[4],"learning,":[5],"continuous":[6],"learning":[7],"and":[8,33,35,84,97,108,113,119,136,155,160,174],"many":[9],"other":[10],"applications,":[11],"on-device":[12,58,195],"training":[13,40],"remains":[14],"an":[15],"open":[16],"problem":[17,21],"EdgeAI.":[19],"The":[20],"stems":[22],"from":[23],"the":[24,42,75,81,117],"large":[25],"number":[26],"of":[27,87,122,188],"operations":[28],"(e.g.,":[29,104,110],"floating":[30],"point":[31],"multiplications":[32],"additions)":[34],"memory":[36,85,138],"consumption":[37,86],"required":[38],"during":[39,90],"by":[41],"back-propagation":[43],"algorithm.":[44],"Consequently,":[45,180],"in":[46,74,172],"this":[47],"paper,":[48],"we":[49,130],"propose":[50],"a":[51,67,185,191],"new":[52,186],"gradient":[53,76],"filtering":[54],"approach":[55,65,182],"which":[56],"enables":[57],"CNN":[59,102],"model":[60],"training.":[61,91,196],"More":[62],"precisely,":[63],"our":[64,123,149,181],"creates":[66],"special":[68],"structure":[69],"with":[70,100,143,190],"fewer":[71],"unique":[72],"elements":[73],"map,":[77],"thus":[78],"significantly":[79],"reducing":[80],"computational":[82],"complexity":[83],"back":[88],"propagation":[89],"Extensive":[92],"experiments":[93],"on":[94,140,176],"image":[95],"classification":[96,142],"semantic":[98],"segmentation":[99],"multiple":[101],"models":[103],"MobileNet,":[105],"DeepLabV3,":[106],"UPerNet)":[107],"devices":[109],"Raspberry":[111],"Pi":[112],"Jetson":[114,178],"Nano)":[115],"demonstrate":[116],"effectiveness":[118],"wide":[120],"applicability":[121],"approach.":[124],"For":[125],"example,":[126],"compared":[127,167],"to":[128,133,153,168],"SOTA,":[129],"achieve":[131],"up":[132,184],"19$\\times$":[134],"speedup":[135,159],"77.1%":[137],"savings":[139,163],"ImageNet":[141],"only":[144],"0.1%":[145],"accuracy":[146],"loss.":[147],"Finally,":[148],"method":[150],"is":[151],"easy":[152],"implement":[154],"deploy;":[156],"over":[157],"20$\\times$":[158],"90%":[161],"energy":[162],"have":[164],"been":[165],"observed":[166],"highly":[169],"optimized":[170],"baselines":[171],"MKLDNN":[173],"CUDNN":[175],"NVIDIA":[177],"Nano.":[179],"opens":[183],"direction":[187],"research":[189],"huge":[192],"potential":[193]},"counts_by_year":[],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2025-10-10T00:00:00"}