{"id":"https://openalex.org/W4387010776","doi":"https://doi.org/10.1109/tpami.2023.3319005","title":"Re-Thinking the Effectiveness of Batch Normalization and Beyond","display_name":"Re-Thinking the Effectiveness of Batch Normalization and Beyond","publication_year":2023,"publication_date":"2023-09-25","ids":{"openalex":"https://openalex.org/W4387010776","doi":"https://doi.org/10.1109/tpami.2023.3319005","pmid":"https://pubmed.ncbi.nlm.nih.gov/37747867"},"language":"en","primary_location":{"id":"doi:10.1109/tpami.2023.3319005","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpami.2023.3319005","pdf_url":null,"source":{"id":"https://openalex.org/S199944782","display_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","issn_l":"0162-8828","issn":["0162-8828","1939-3539","2160-9292"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320439","host_organization_name":"IEEE Computer Society","host_organization_lineage":["https://openalex.org/P4310320439","https://openalex.org/P4310319808"],"host_organization_lineage_names":["IEEE Computer Society","Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"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/A5006711084","display_name":"Hanyang Peng","orcid":"https://orcid.org/0000-0002-9715-473X"},"institutions":[{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Hanyang Peng","raw_affiliation_strings":["Peng Cheng Laboratory, Shenzhen, China"],"affiliations":[{"raw_affiliation_string":"Peng Cheng Laboratory, Shenzhen, China","institution_ids":["https://openalex.org/I4210136793"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100397991","display_name":"Yue Yu","orcid":"https://orcid.org/0000-0002-9865-2212"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]},{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yue Yu","raw_affiliation_strings":["Peng Cheng Laboratory, Shenzhen, China","College of Computer, National University of Defense Technology, Changsha, China"],"affiliations":[{"raw_affiliation_string":"Peng Cheng Laboratory, Shenzhen, China","institution_ids":["https://openalex.org/I4210136793"]},{"raw_affiliation_string":"College of Computer, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5030831939","display_name":"Shiqi Yu","orcid":"https://orcid.org/0000-0002-5213-5877"},"institutions":[{"id":"https://openalex.org/I3045169105","display_name":"Southern University of Science and Technology","ror":"https://ror.org/049tv2d57","country_code":"CN","type":"education","lineage":["https://openalex.org/I3045169105"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shiqi Yu","raw_affiliation_strings":["Department of Computer Science and Engineering, Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology, Shenzhen, China"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology, Shenzhen, China","institution_ids":["https://openalex.org/I3045169105"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5006711084"],"corresponding_institution_ids":["https://openalex.org/I4210136793"],"apc_list":null,"apc_paid":null,"fwci":5.9148,"has_fulltext":false,"cited_by_count":34,"citation_normalized_percentile":{"value":0.97060066,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":"46","issue":"1","first_page":"465","last_page":"478"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","score":0.998199999332428,"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/T12676","display_name":"Machine Learning and ELM","score":0.998199999332428,"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":0.996999979019165,"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/T12702","display_name":"Brain Tumor Detection and Classification","score":0.9941999912261963,"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/normalization","display_name":"Normalization (sociology)","score":0.7535319924354553},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.7049720287322998},{"id":"https://openalex.org/keywords/variance-reduction","display_name":"Variance reduction","score":0.6112591624259949},{"id":"https://openalex.org/keywords/lipschitz-continuity","display_name":"Lipschitz continuity","score":0.574481725692749},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.52494215965271},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.4686870872974396},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.45539870858192444},{"id":"https://openalex.org/keywords/sigmoid-function","display_name":"Sigmoid function","score":0.43557512760162354},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.40751805901527405},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.40649405121803284},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.3561575412750244},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.34871768951416016},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.34021586179733276},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.1630449891090393}],"concepts":[{"id":"https://openalex.org/C136886441","wikidata":"https://www.wikidata.org/wiki/Q926129","display_name":"Normalization (sociology)","level":2,"score":0.7535319924354553},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.7049720287322998},{"id":"https://openalex.org/C62644790","wikidata":"https://www.wikidata.org/wiki/Q3454689","display_name":"Variance reduction","level":3,"score":0.6112591624259949},{"id":"https://openalex.org/C22324862","wikidata":"https://www.wikidata.org/wiki/Q652707","display_name":"Lipschitz continuity","level":2,"score":0.574481725692749},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.52494215965271},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.4686870872974396},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.45539870858192444},{"id":"https://openalex.org/C81388566","wikidata":"https://www.wikidata.org/wiki/Q526668","display_name":"Sigmoid function","level":3,"score":0.43557512760162354},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.40751805901527405},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40649405121803284},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3561575412750244},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.34871768951416016},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.34021586179733276},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.1630449891090393},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C19499675","wikidata":"https://www.wikidata.org/wiki/Q232207","display_name":"Monte Carlo method","level":2,"score":0.0},{"id":"https://openalex.org/C144024400","wikidata":"https://www.wikidata.org/wiki/Q21201","display_name":"Sociology","level":0,"score":0.0},{"id":"https://openalex.org/C19165224","wikidata":"https://www.wikidata.org/wiki/Q23404","display_name":"Anthropology","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/tpami.2023.3319005","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpami.2023.3319005","pdf_url":null,"source":{"id":"https://openalex.org/S199944782","display_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","issn_l":"0162-8828","issn":["0162-8828","1939-3539","2160-9292"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320439","host_organization_name":"IEEE Computer Society","host_organization_lineage":["https://openalex.org/P4310320439","https://openalex.org/P4310319808"],"host_organization_lineage_names":["IEEE Computer Society","Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","raw_type":"journal-article"},{"id":"pmid:37747867","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/37747867","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on pattern analysis and machine intelligence","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3309959654","display_name":null,"funder_award_id":"61976144","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G8914810884","display_name":null,"funder_award_id":"61806128","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":55,"referenced_works":["https://openalex.org/W6908809","https://openalex.org/W1686810756","https://openalex.org/W2194775991","https://openalex.org/W2502312327","https://openalex.org/W2785678896","https://openalex.org/W2787911558","https://openalex.org/W2793888044","https://openalex.org/W2904614653","https://openalex.org/W2927966956","https://openalex.org/W2963016543","https://openalex.org/W2963743626","https://openalex.org/W2997209116","https://openalex.org/W3035524453","https://openalex.org/W3110258731","https://openalex.org/W3172624760","https://openalex.org/W3209259379","https://openalex.org/W4250482878","https://openalex.org/W4293455017","https://openalex.org/W4302375066","https://openalex.org/W6600284362","https://openalex.org/W6631190155","https://openalex.org/W6638667902","https://openalex.org/W6676105031","https://openalex.org/W6685315510","https://openalex.org/W6695676441","https://openalex.org/W6698102009","https://openalex.org/W6701650085","https://openalex.org/W6724804524","https://openalex.org/W6733590821","https://openalex.org/W6739901393","https://openalex.org/W6747620207","https://openalex.org/W6748019269","https://openalex.org/W6748582592","https://openalex.org/W6751923770","https://openalex.org/W6752314195","https://openalex.org/W6752495264","https://openalex.org/W6753122105","https://openalex.org/W6753892653","https://openalex.org/W6755144094","https://openalex.org/W6756001544","https://openalex.org/W6757276028","https://openalex.org/W6758094429","https://openalex.org/W6758207765","https://openalex.org/W6762718338","https://openalex.org/W6762875896","https://openalex.org/W6763249885","https://openalex.org/W6764442487","https://openalex.org/W6767560732","https://openalex.org/W6768511045","https://openalex.org/W6771763659","https://openalex.org/W6780226713","https://openalex.org/W6785036545","https://openalex.org/W6785827884","https://openalex.org/W6797288013","https://openalex.org/W6804657168"],"related_works":["https://openalex.org/W2107438106","https://openalex.org/W4288346661","https://openalex.org/W2947416323","https://openalex.org/W2895097035","https://openalex.org/W4206903459","https://openalex.org/W2754816816","https://openalex.org/W4283750944","https://openalex.org/W4366280654","https://openalex.org/W3160167280","https://openalex.org/W3020853991"],"abstract_inverted_index":{"Batch":[0],"normalization":[1,179,184,192],"(BN)":[2],"is":[3,33,57,68,90,205,298],"used":[4],"by":[5,287],"default":[6],"in":[7,17,167],"many":[8],"modern":[9],"deep":[10],"neural":[11],"networks":[12,255,279],"due":[13,34],"to":[14,35,59,73,92,103,244,301],"its":[15],"effectiveness":[16,30,62],"accelerating":[18],"training":[19,236,241],"convergence":[20,102,169,237],"and":[21,41,65,106,127,163,193,229,239,264,283,292],"boosting":[22],"inference":[23],"performance.":[24,170],"Recent":[25],"studies":[26],"suggest":[27],"that":[28,84,139,207,232],"the":[29,36,39,45,61,111,118,121,124,131,134,147,188,195,202,212,218,235,240,275,302],"of":[31,38,47,63,110,117,120,123,130,133,191,197,217],"BN":[32,64,72,141,156,199],"Lipschitzness":[37,56],"loss":[40,242],"gradient,":[42,126],"rather":[43,151],"than":[44,152,249,289],"reduction":[46,109,116,129],"internal":[48],"covariate":[49],"shift.":[50],"However,":[51],"questions":[52],"remain":[53],"about":[54],"whether":[55,66],"sufficient":[58],"explain":[60],"there":[67],"room":[69],"for":[70,278,304],"vanilla":[71,140,155,173,198,250],"be":[74,104],"further":[75],"improved.":[76],"To":[77,171],"answer":[78],"these":[79],"questions,":[80],"we":[81,175],"first":[82],"prove":[83],"when":[85],"stochastic":[86,125,135],"gradient":[87,112],"descent":[88],"(SGD)":[89],"applied":[91],"optimize":[93],"a":[94,177,245],"general":[95],"non-convex":[96],"problem,":[97],"three":[98,148,213],"effects":[99,149,214],"will":[100,165],"help":[101],"faster":[105],"better:":[107],"(i)":[108],"Lipschitz":[113],"constant,":[114],"(ii)":[115],"expectation":[119],"square":[122],"(iii)":[128],"variance":[132],"gradient.":[136],"We":[137],"demonstrate":[138],"only":[142],"with":[143,157,256,280],"ReLU":[144],"can":[145,209],"induce":[146],"above,":[150,215],"Lipschitzness,":[153],"but":[154],"other":[158],"nonlinearities":[159],"like":[160],"Sigmoid,":[161,281],"Tanh,":[162,261,282],"SELU":[164,284],"result":[166],"degraded":[168],"improve":[172],"BN,":[174],"propose":[176],"new":[178],"approach,":[180],"dubbed":[181],"complete":[182],"batch":[183],"(CBN),":[185],"which":[186,297],"changes":[187],"placement":[189],"position":[190],"modifies":[194],"structure":[196],"based":[200],"on":[201,224,294],"theory.":[203],"It":[204],"proven":[206],"CBN":[208,233,253],"elicit":[210],"all":[211],"regardless":[216],"nonlinear":[219,258],"activation":[220],"used.":[221],"Extensive":[222],"experiments":[223],"benchmark":[225],"datasets":[226],"CIFAR10,":[227],"CIFAR100,":[228],"ILSVRC2012":[230],"validate":[231],"makes":[234],"faster,":[238],"converges":[243],"smaller":[246],"local":[247],"minimum":[248],"BN.":[251],"Moreover,":[252],"helps":[254],"multiple":[257],"activations":[259],"(Sigmoid,":[260],"ReLU,":[262],"SELU,":[263],"Swish)":[265],"achieve":[266],"higher":[267],"test":[268],"accuracy":[269],"steadily.":[270],"Specifically,":[271],"benefitting":[272],"from":[273],"CBN,":[274],"classification":[276],"accuracies":[277],"are":[285],"boosted":[286],"more":[288],"15.0%,":[290],"4.5%,":[291],"4.0%":[293],"average,":[295],"respectively,":[296],"even":[299],"comparable":[300],"performance":[303],"ReLU.":[305]},"counts_by_year":[{"year":2026,"cited_by_count":4},{"year":2025,"cited_by_count":19},{"year":2024,"cited_by_count":11}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}