{"id":"https://openalex.org/W29834869","doi":"https://doi.org/10.21437/eurospeech.1997-406","title":"A new algorithm for robust speech recognition: the delta vector taylor series approach","display_name":"A new algorithm for robust speech recognition: the delta vector taylor series approach","publication_year":1997,"publication_date":"1997-09-22","ids":{"openalex":"https://openalex.org/W29834869","doi":"https://doi.org/10.21437/eurospeech.1997-406","mag":"29834869"},"language":"en","primary_location":{"id":"doi:10.21437/eurospeech.1997-406","is_oa":false,"landing_page_url":"https://doi.org/10.21437/eurospeech.1997-406","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"5th European Conference on Speech Communication and Technology (Eurospeech 1997)","raw_type":"proceedings-article"},"type":"article","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/A5103874391","display_name":"Pedro J. Moreno","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Pedro J. Moreno","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5061680523","display_name":"Brian Eberman","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Brian Eberman","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5103874391"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.0037594,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"2599","last_page":"2602"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9994999766349792,"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"}},"topics":[{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9994999766349792,"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"}},{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9991999864578247,"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/T11447","display_name":"Blind Source Separation Techniques","score":0.994700014591217,"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/algorithm","display_name":"Algorithm","score":0.6943805813789368},{"id":"https://openalex.org/keywords/timit","display_name":"TIMIT","score":0.605177104473114},{"id":"https://openalex.org/keywords/series","display_name":"Series (stratigraphy)","score":0.5854395627975464},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5415137410163879},{"id":"https://openalex.org/keywords/taylor-series","display_name":"Taylor series","score":0.4893304705619812},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.4578496515750885},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4179186224937439},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.4045935869216919},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.26730749011039734},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.21031233668327332}],"concepts":[{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.6943805813789368},{"id":"https://openalex.org/C2778724510","wikidata":"https://www.wikidata.org/wiki/Q7670405","display_name":"TIMIT","level":3,"score":0.605177104473114},{"id":"https://openalex.org/C143724316","wikidata":"https://www.wikidata.org/wiki/Q312468","display_name":"Series (stratigraphy)","level":2,"score":0.5854395627975464},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5415137410163879},{"id":"https://openalex.org/C158946198","wikidata":"https://www.wikidata.org/wiki/Q131187","display_name":"Taylor series","level":2,"score":0.4893304705619812},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4578496515750885},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4179186224937439},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.4045935869216919},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.26730749011039734},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.21031233668327332},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"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/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/eurospeech.1997-406","is_oa":false,"landing_page_url":"https://doi.org/10.21437/eurospeech.1997-406","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"5th European Conference on Speech Communication and Technology (Eurospeech 1997)","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":5,"referenced_works":["https://openalex.org/W105750483","https://openalex.org/W1640154593","https://openalex.org/W1665196592","https://openalex.org/W2157590573","https://openalex.org/W4285719527"],"related_works":["https://openalex.org/W2071828724","https://openalex.org/W4401950215","https://openalex.org/W1994694193","https://openalex.org/W2161865195","https://openalex.org/W170831052","https://openalex.org/W4247725880","https://openalex.org/W1006721676","https://openalex.org/W2100854157","https://openalex.org/W1771058801","https://openalex.org/W2978471304"],"abstract_inverted_index":{"A":[0],"New":[1],"Algorithm":[2],"for":[3,65,100,384],"Robust":[4],"Sp":[5],"eech":[6,70,159,208,255,292,387,403,462,547],"Recognition:":[7],"The":[8,103,178,485],"Delta":[9,36,359],"VectorTaylor":[10],"Series":[11,39,50,362],"ApproachPedroJ.Moreno":[12],"and":[13,46,150,167,227,238,244,262,286,346,353,380,448,525,554],"Brian":[14],"Ebermanemail:":[15],"pjm@crl.dec.com,":[16],"bse@crl.deomDigital":[17],"Equipment":[18],"Corp":[19],"orationCambridge":[20],"Research":[21],"Lab":[22],"oratoryABSTRACTIn":[23],"this":[24,367,535,600],"pap":[25,409,536],"er":[26,410,537],"we":[27,61,424,442,453,538],"present":[28],"a":[29,75,108,125,140,153,268,311,342,376,381,396,438,466,483,540],"new":[30,126],"mo":[31,242,279,429,458,476,541],"del-based":[32],"comp":[33,127,245],"ensationtechnique":[34],"called":[35],"Vector":[37,360],"Taylor":[38,49,361],"(DVTS).":[40],"Thisnew":[41],"technique":[42,336],"is":[43,105,170,411,475,548,569,586],"an":[44],"extension":[45],"improvementoer":[47],"theVector":[48],"(VTS)":[51],"approach":[52,129],"[7]":[53,517],"that":[54],"addressesseveral":[55],"of":[56,68,134,206,214,250,271,310,322,341,407,468,481,487,498,505,514,528,543,564,575,582,592],"its":[57,228,330,356],"limitations":[58],".In":[59],"particular,":[60],"presentanew":[62],"statistical":[63,313,344,378,399],"representation":[64,314,345,379,497],"the":[66,85,113,132,146,192,204,224,247,251,281,289,298,306,320,324,339,348,372,401,408,419,427,433,460,499,511,515,529,544,559,565,570,576,579,583,587,593,596],"distribution":[67],"cleansp":[69,577],"feature":[71,404,463],"vectors":[72,285,464],"based":[73,130],"on":[74,131,144,253],"weighted":[76],"vector":[77],"co":[78],"de-b":[79],"o":[80],"ok.":[81],"This":[82],"change":[83],"to":[84,93,241,278,294,370,391,432,494],"underlying":[86],"probabili":[87,489],"ty":[88],"densityfunction":[89],"(PDF)":[90],"allows":[91],"us":[92],"pro":[94,520],"duce":[95],"more":[96,613],"accurate":[97],"andstable":[98],"solutions":[99],"our":[101,137],"algorithm.":[102,518],"algorithm":[104,138,179,434],"alsopresented":[106],"in":[107,139,152,210,235,355,366,450,533,612],"EM-MAP":[109],"framework":[110],"where":[111],"some":[112],"en-vironmental":[114],"parameters":[115],"are":[116],"treated":[117],"as":[118,217,412,437,465],"random":[119],"variableswith":[120],"known":[121],"PDF's.":[122],"Finally,we":[123],"explore":[124],"ensa-tion":[128],"use":[133,309],"convex":[135],"hulls.Weevaluate":[136],"phonetic":[141],"classi":[142],"cati":[143],"taskon":[145],"TIMIT":[147],"[5]":[148],"database":[149],"also":[151],"small":[154],"vo":[155],"cabu-lary":[156],"size":[157],"sp":[158,207,254,283,291,326,386,402,461,546,562,567,573,590],"recognition":[160,209,225,307,335],"database.":[161],"In":[162,414],"b":[163,198,300,302,350,373,602],"oth":[164],"databasesarti":[165],"cial":[166],"natural":[168],"noise":[169],"injected":[171],"at":[172,184,347],"several":[173,195,502],"signal":[174],"tonoise":[175],"ratios":[176],"(SNR).":[177],"achieves":[180],"matched":[181],"p":[182,316,560,571,588],"er-formance":[183],"all":[185,506],"SNR's":[186],"ab":[187],"ove":[188],"10":[189],"dB.1.Intro":[190],"ductionOver":[191],"last":[193],"years":[194],"techniques":[196,258],"have":[197,266,338],"een":[199],"prop":[200],"osedto":[201],"deal":[202],"with":[203,329],"problem":[205],"noisy":[211,290],"en-vironments.":[212],"Some":[213],"them":[215],"such":[216],"PMC":[218],"[3],":[219],"or":[220,276],"MLLR":[221],"[6]have":[222],"used":[223,267],"engine":[226,328],"rich":[229,312,343,377],"statisticalrepresentation":[230],"(more":[231],"than":[232],"90,000":[233],"Gaussians":[234],"systemslike":[236],"SPHINX-3":[237],"HTK":[239],"[9])":[240],"del":[243,280,542],"ensatefor":[246],"e":[248,295,418,426,444,603],"ects":[249],"environment":[252],"recognitionsystems.":[256],"Other":[257],"like":[259],"CDCN":[260],"[1]":[261],"POF":[263],"[8]among":[264],"others":[265],"reduced":[269],"set":[270],"Gaussian":[272],"mix-tures":[273],"(typically":[274],"256":[275],"less)":[277],"clean":[297],"eechfeature":[284],"prepro":[287],"cess":[288],"featuresvectors":[293],"ectively":[296],"features":[299],"efore":[301],"eing":[303,351],"pro-cessed":[304],"by":[305,394,550],"engine.The":[308],"improves":[315],"er-formance,":[317],"but":[318],"has":[319],"drawback":[321],"using":[323,395],"whole":[325],"eechrecognition":[327],"asso":[331],"ciated":[332],"complexity.Anideal":[333],"robust":[334,385,524],"should":[337],"advan-tages":[340],"sametime":[349],"simple":[352],"fast":[354],"op":[357],"eration.The":[358],"(DVTS)":[363],"approachis":[364],"anattempt":[365],"direction.It":[368],"tries":[369,390],"gain":[371],"ene":[374],"tsof":[375],"low":[382],"complexitytechnique":[383],"recognition.":[388],"It":[389,519],"achievethese":[392],"goals":[393],"di":[397,430],"erent":[398],"representationfor":[400],"vectors.The":[405],"outline":[406],"follows.":[413],"section":[415,422,440,451],"2":[416],"wedescrib":[417],"DVTS":[420],"algorithm.In":[421],"3":[423],"brieydescrib":[425],"necessary":[428],"cations":[431],"tomakeit":[435],"work":[436],"lter.In":[439],"4":[441],"describ":[443],"ourexp":[445],"erimental":[446],"results":[447],"nally":[449],"5":[452],"presentour":[454],"conclusions.2.New":[455],"Algorithm:":[456],"Delta-VTSDVTS":[457],"dels":[459],"weightedsum":[467],"multidimensi":[469],"onal":[470],"Dirac":[471],"deltasp(x)=M1Xk=0P[k]\u000e)(1)where":[472],"eachvector":[473],"function\u000e(xk)":[474],"deled":[477],"as\u000e(xk)=D1Yi=0ii;k)(2)P[k]is":[478],"ana":[479],"prioriprobability":[480],"observing":[482],"particulardelta.":[484],"sum":[486],"these":[488],"ties":[490],"must":[491],"add":[492],"up":[493],"one.This":[495],"novel":[496],"PDF":[500],"ofxhas":[501],"ad-vantages.":[503],"First":[504],"it":[507],"greatly":[508],"simpli":[509],"es":[510],"mathemat-ical":[512],"assumptions":[513],"VTS":[516],"ducesa":[521],"simple,":[522],"fast,":[523],"direct":[526],"formulation":[527],"EMsolutions":[530],"already":[531],"presented":[532],"[7].In":[534],"assume":[539],"environmentinwhich":[545],"corrupted":[549],"unknown":[551,555],"additive":[552,594],"stationarynoise":[553],"linear":[556,584],"lteringZ(!)=X)jH2+N(3)whereZ(!)":[557],"represents":[558],"ower":[561,572,589],"ectrum":[563,574,591],"de-graded":[566],"eech,X(!)":[568],"eech,jH(!)2is":[578],"transfer":[580],"function":[581],"lter,andN(!)":[585],"noise.In":[595],"log-mel-sp":[597],"ectral":[598],"domain":[599],"can":[601],"expressed":[604],"asz=x+":[605],"log":[606],"(exp":[607],"(q)":[608],"+":[609],"exp":[610],"(n))(4)or":[611],"general":[614],"termsz=x+f(;nq)(5)":[615]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}