{"id":"https://openalex.org/W3089425003","doi":"https://doi.org/10.1109/ijcnn48605.2020.9207145","title":"Robust Training of Vector Quantized Bottleneck Models","display_name":"Robust Training of Vector Quantized Bottleneck Models","publication_year":2020,"publication_date":"2020-07-01","ids":{"openalex":"https://openalex.org/W3089425003","doi":"https://doi.org/10.1109/ijcnn48605.2020.9207145","mag":"3089425003"},"language":"en","primary_location":{"id":"doi:10.1109/ijcnn48605.2020.9207145","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ijcnn48605.2020.9207145","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2020 International Joint Conference on Neural Networks (IJCNN)","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/A5011669714","display_name":"Adrian \u0141a\u0144cucki","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Adrian Lancucki","raw_affiliation_strings":["NVIDIA Corporation, Warsaw, Poland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA Corporation, Warsaw, Poland","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000408865","display_name":"Jan Chorowski","orcid":"https://orcid.org/0000-0002-1570-7610"},"institutions":[{"id":"https://openalex.org/I219388962","display_name":"University of Wroc\u0142aw","ror":"https://ror.org/00yae6e25","country_code":"PL","type":"education","lineage":["https://openalex.org/I219388962"]}],"countries":["PL"],"is_corresponding":false,"raw_author_name":"Jan Chorowski","raw_affiliation_strings":["University of Wroclaw, Wroclaw, Poland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Wroclaw, Wroclaw, Poland","institution_ids":["https://openalex.org/I219388962"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006277764","display_name":"Guillaume Sanchez","orcid":null},"institutions":[{"id":"https://openalex.org/I143002897","display_name":"Universit\u00e9 de Toulon","ror":"https://ror.org/02m9kbe37","country_code":"FR","type":"education","lineage":["https://openalex.org/I143002897"]},{"id":"https://openalex.org/I4210114274","display_name":"Laboratoire d\u2019Informatique et Syst\u00e8mes","ror":"https://ror.org/0257sgk90","country_code":"FR","type":"facility","lineage":["https://openalex.org/I1294671590","https://openalex.org/I143002897","https://openalex.org/I21491767","https://openalex.org/I4210114274"]}],"countries":["FR"],"is_corresponding":false,"raw_author_name":"Guillaume Sanchez","raw_affiliation_strings":["LIS, Universit\u00e9 de Toulon, Toulon, France"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LIS, Universit\u00e9 de Toulon, Toulon, France","institution_ids":["https://openalex.org/I143002897","https://openalex.org/I4210114274"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013266148","display_name":"Ricard Marxer","orcid":"https://orcid.org/0000-0001-5099-5059"},"institutions":[{"id":"https://openalex.org/I143002897","display_name":"Universit\u00e9 de Toulon","ror":"https://ror.org/02m9kbe37","country_code":"FR","type":"education","lineage":["https://openalex.org/I143002897"]},{"id":"https://openalex.org/I4210114274","display_name":"Laboratoire d\u2019Informatique et Syst\u00e8mes","ror":"https://ror.org/0257sgk90","country_code":"FR","type":"facility","lineage":["https://openalex.org/I1294671590","https://openalex.org/I143002897","https://openalex.org/I21491767","https://openalex.org/I4210114274"]}],"countries":["FR"],"is_corresponding":false,"raw_author_name":"Ricard Marxer","raw_affiliation_strings":["LIS, Universit\u00e9 de Toulon, Toulon, France"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LIS, Universit\u00e9 de Toulon, Toulon, France","institution_ids":["https://openalex.org/I143002897","https://openalex.org/I4210114274"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071702566","display_name":"Nanxin Chen","orcid":"https://orcid.org/0000-0001-6698-1604"},"institutions":[{"id":"https://openalex.org/I145311948","display_name":"Johns Hopkins University","ror":"https://ror.org/00za53h95","country_code":"US","type":"education","lineage":["https://openalex.org/I145311948"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nanxin Chen","raw_affiliation_strings":["Johns Hopkins University, Baltimore, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Johns Hopkins University, Baltimore, USA","institution_ids":["https://openalex.org/I145311948"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001650589","display_name":"Hans J. G. A. Dolfing","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hans J.G.A. Dolfing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035187454","display_name":"Sameer Khurana","orcid":"https://orcid.org/0000-0002-3182-1085"},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sameer Khurana","raw_affiliation_strings":["Massachusetts Institute of Technology, Cambridge, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Massachusetts Institute of Technology, Cambridge, USA","institution_ids":["https://openalex.org/I63966007"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023201445","display_name":"Tanel Alum\u00e4e","orcid":"https://orcid.org/0000-0001-5083-1556"},"institutions":[{"id":"https://openalex.org/I111112146","display_name":"Tallinn University of Technology","ror":"https://ror.org/0443cwa12","country_code":"EE","type":"education","lineage":["https://openalex.org/I111112146"]}],"countries":["EE"],"is_corresponding":false,"raw_author_name":"Tanel Alumae","raw_affiliation_strings":["Tallinn University of Technology, Tallinn, Estonia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tallinn University of Technology, Tallinn, Estonia","institution_ids":["https://openalex.org/I111112146"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5025212823","display_name":"Antoine Laurent","orcid":"https://orcid.org/0000-0002-2653-1008"},"institutions":[{"id":"https://openalex.org/I4210108471","display_name":"Le Mans Universit\u00e9","ror":"https://ror.org/01mtcc283","country_code":"FR","type":"education","lineage":["https://openalex.org/I4210108471"]}],"countries":["FR"],"is_corresponding":false,"raw_author_name":"Antoine Laurent","raw_affiliation_strings":["Le Mans University, Le Mans, France"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Le Mans University, Le Mans, France","institution_ids":["https://openalex.org/I4210108471"]}]}],"institutions":[],"countries_distinct_count":4,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":35,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"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/T10028","display_name":"Topic Modeling","score":0.9987999796867371,"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/T11309","display_name":"Music and Audio Processing","score":0.9986000061035156,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/codebook","display_name":"Codebook","score":0.9058411717414856},{"id":"https://openalex.org/keywords/initialization","display_name":"Initialization","score":0.7791236639022827},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6709816455841064},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.5905523300170898},{"id":"https://openalex.org/keywords/latent-variable","display_name":"Latent variable","score":0.5808295011520386},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.5629774928092957},{"id":"https://openalex.org/keywords/autoencoder","display_name":"Autoencoder","score":0.5613430142402649},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5609806776046753},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.49702003598213196},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.4862147867679596},{"id":"https://openalex.org/keywords/unsupervised-learning","display_name":"Unsupervised learning","score":0.4595903158187866},{"id":"https://openalex.org/keywords/information-bottleneck-method","display_name":"Information bottleneck method","score":0.4418639838695526},{"id":"https://openalex.org/keywords/code-word","display_name":"Code word","score":0.44013673067092896},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.4329095482826233},{"id":"https://openalex.org/keywords/feature-learning","display_name":"Feature learning","score":0.41434380412101746},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.33365398645401},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3187936544418335},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.20795106887817383}],"concepts":[{"id":"https://openalex.org/C127759330","wikidata":"https://www.wikidata.org/wiki/Q637416","display_name":"Codebook","level":2,"score":0.9058411717414856},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.7791236639022827},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6709816455841064},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.5905523300170898},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.5808295011520386},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.5629774928092957},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.5613430142402649},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5609806776046753},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.49702003598213196},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4862147867679596},{"id":"https://openalex.org/C8038995","wikidata":"https://www.wikidata.org/wiki/Q1152135","display_name":"Unsupervised learning","level":2,"score":0.4595903158187866},{"id":"https://openalex.org/C60008888","wikidata":"https://www.wikidata.org/wiki/Q6031013","display_name":"Information bottleneck method","level":3,"score":0.4418639838695526},{"id":"https://openalex.org/C153207627","wikidata":"https://www.wikidata.org/wiki/Q863873","display_name":"Code word","level":3,"score":0.44013673067092896},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.4329095482826233},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.41434380412101746},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.33365398645401},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3187936544418335},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.20795106887817383},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C94625758","wikidata":"https://www.wikidata.org/wiki/Q7163","display_name":"Politics","level":2,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ijcnn48605.2020.9207145","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ijcnn48605.2020.9207145","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2020 International Joint Conference on Neural Networks (IJCNN)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":66,"referenced_works":["https://openalex.org/W1522301498","https://openalex.org/W1524333225","https://openalex.org/W1583776211","https://openalex.org/W1686946872","https://openalex.org/W1836465849","https://openalex.org/W1902027874","https://openalex.org/W1959608418","https://openalex.org/W1970890968","https://openalex.org/W2073459066","https://openalex.org/W2086161653","https://openalex.org/W2100768664","https://openalex.org/W2119717200","https://openalex.org/W2119885577","https://openalex.org/W2123237149","https://openalex.org/W2127141656","https://openalex.org/W2145889472","https://openalex.org/W2193413348","https://openalex.org/W2267126114","https://openalex.org/W2293634267","https://openalex.org/W2347098582","https://openalex.org/W2464234964","https://openalex.org/W2547875792","https://openalex.org/W2557579533","https://openalex.org/W2602076750","https://openalex.org/W2750248772","https://openalex.org/W2752796333","https://openalex.org/W2804145368","https://openalex.org/W2887927938","https://openalex.org/W2888911345","https://openalex.org/W2940544976","https://openalex.org/W2947590261","https://openalex.org/W2949117887","https://openalex.org/W2951004968","https://openalex.org/W2953318193","https://openalex.org/W2962695963","https://openalex.org/W2963799213","https://openalex.org/W2964121744","https://openalex.org/W2971074500","https://openalex.org/W2972374322","https://openalex.org/W2972867623","https://openalex.org/W2973026522","https://openalex.org/W2979454998","https://openalex.org/W3008499099","https://openalex.org/W3049315473","https://openalex.org/W3100270690","https://openalex.org/W3125709657","https://openalex.org/W3136591341","https://openalex.org/W4293469690","https://openalex.org/W6631190155","https://openalex.org/W6631362777","https://openalex.org/W6634851332","https://openalex.org/W6637108112","https://openalex.org/W6638667902","https://openalex.org/W6640963894","https://openalex.org/W6668990524","https://openalex.org/W6675022971","https://openalex.org/W6687566353","https://openalex.org/W6696934422","https://openalex.org/W6719357382","https://openalex.org/W6729448088","https://openalex.org/W6729906282","https://openalex.org/W6735799286","https://openalex.org/W6751821244","https://openalex.org/W6753966407","https://openalex.org/W6762931180","https://openalex.org/W6774456908"],"related_works":["https://openalex.org/W2369939250","https://openalex.org/W2355032504","https://openalex.org/W1588525449","https://openalex.org/W2050865889","https://openalex.org/W2375798515","https://openalex.org/W2145406111","https://openalex.org/W1966593131","https://openalex.org/W2393018242","https://openalex.org/W4376277154","https://openalex.org/W4384933762"],"abstract_inverted_index":{"In":[0,82],"this":[1,83],"paper":[2,84],"we":[3,85,152],"demonstrate":[4,129],"methods":[5],"for":[6,141,169,176],"reliable":[7],"and":[8,38,122,144,160],"efficient":[9],"training":[10,66,156],"of":[11,31,78,124,165,189],"discrete":[12,68,91],"representation":[13,48,180],"using":[14],"Vector-Quantized":[15],"Variational":[16,62],"Auto-Encoder":[17,63],"models":[18,23,58,70],"(VQ-VAEs).":[19],"Discrete":[20],"latent":[21,56,166,190],"variable":[22,57,69],"have":[24],"been":[25],"shown":[26,94],"to":[27,34,54,74,95,187],"learn":[28],"nontrivial":[29],"representations":[30],"speech,":[32],"applicable":[33],"unsupervised":[35,47,179],"voice":[36],"conversion":[37],"reaching":[39],"state-of-the-art":[40,90],"performance":[41],"on":[42,87,97],"unit":[43],"discovery":[44],"tasks.":[45],"For":[46],"learning,":[49,181],"they":[50],"became":[51],"viable":[52],"alternatives":[53],"continuous":[55,101],"such":[59],"as":[60],"the":[61,75,79,114,138,142,163],"(VAE).":[64],"However,":[65],"deep":[67],"is":[71],"challenging,":[72],"due":[73],"inherent":[76],"non-differentiability":[77],"discretization":[80],"operation.":[81],"focus":[86],"VQ-VAE,":[88],"a":[89,150],"bottleneck":[92],"model":[93],"perform":[96],"par":[98],"with":[99,107],"its":[100],"counterparts.":[102],"It":[103],"quantizes":[104],"encoder":[105,126],"outputs":[106],"on-line":[108],"k-means":[109],"clustering.":[110],"We":[111,128],"show":[112],"that":[113,130],"codebook":[115,143],"learning":[116,139],"can":[117,132],"suffer":[118],"from":[119],"poor":[120],"initialization":[121],"non-stationarity":[123],"clustered":[125],"outputs.":[127],"these":[131],"be":[133],"successfully":[134],"overcome":[135],"by":[136],"increasing":[137],"rate":[140],"periodic":[145],"date-dependent":[146],"codeword":[147],"re-initialization.":[148],"As":[149],"result,":[151],"achieve":[153],"more":[154],"robust":[155],"across":[157],"different":[158],"tasks,":[159],"significantly":[161],"increase":[162],"usage":[164],"codewords":[167],"even":[168],"large":[170,183],"codebooks.":[171],"This":[172],"has":[173],"practical":[174],"benefit,":[175],"instance,":[177],"in":[178],"where":[182],"codebooks":[184],"may":[185],"lead":[186],"disentanglement":[188],"representations.":[191]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":10},{"year":2024,"cited_by_count":8},{"year":2023,"cited_by_count":9},{"year":2022,"cited_by_count":5},{"year":2021,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2020-10-08T00:00:00"}