欢迎来到天天文库
浏览记录
ID:33529622
大小:118.78 KB
页数:18页
时间:2019-02-26
《基于改进矢量量化k-均值lbg算法的语音识别外文文献翻译》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库。
1、毕业设计(论文)外文翻译外文题目:SpeechRecognitionUsingVectorQuantizationthroughModifiedK-meansLBGAlgorithm译文题目:基于改进矢量量化K-均值LBG算法的语音识别文献出处:《ComputerEngineeringandTntelligentSystems》,2012,7(3)外文作者:BalwantA.Sonkdinble,DharmpalDoye字数统计:英文2389单词,13087字符;中文3968汉字外文文献:SpeechRecognitionUsingVectorQuantization
2、throughModifiedK-meansLBGAlgorithmAbstractIntheVectorQuantization,themaintaskistogenerateagoodcodebook.Thedistortionmeasurebetweentheoriginalpatternandthereconstructedpatternshouldbeminimum.Inthispaper,aproposedalgorithmcalledModifiedK・meansLBGalgorithmusedtoobtainagoodcodebook.Thesyste
3、mhasshowngoodperformanceonlimitedvocabularytasks.Keywords:K・meansalgorithm,LBGalgorithm,VectorQuantization,SpeechRecognitionlelntroductionThenaturalwayofcommunicationamonghumanbeingsisthroughspeech.Manyhumanbeingsareexchangingtheinformationthroughmobilephonesaswellasothercommunicationto
4、olsinarealmanner[L・R・Rabineretal.,1993].TheVectorQuantization(VQ)isthefundamentalandmostsuccessfultechniqueusedinspeechcoding,imagecoding,speechrecognition,andspeechsynthesisandspeakerrecognition[S.Furui,1986].Thesetechniquesareappliedfirstlyintheanalysisofspeechwherethemappingoflargeve
5、ctorspaceintoafinitenumberofregionsinthatspace.TheVQtechniquesarecommonlyappliedtodevelopdiscreteorsemi-continuousHMMbasedspeechrecognitionsystem.InVQ,anorderedsetofsignalsamplesorparameterscanbeefficientlycodedbymatchingtheinputvectortoasimilarpatternorcodevector(codeword)inapredefined
6、codebook[[Tzu-ChuenLuetal.,2010].TheVQtechniquesarealsoknownasdataclusteringmethodsinvariousdisciplines.Itisanunsupervisedlearningprocedurewidelyusedinmanyapplications.Thedataclusteringmethodsareclassifiedashardandsoftclusteringmethods.Thesearecentroid-basedparametricclusteringtechnique
7、sbasedonalargeclassofdistortionfunctionsknownasBregmandivergences[ArindamBaneijeeetal.,2005].Inthehardclustering,eachdatapointbelongstoexactlyoneofthepartitionsinobtainingthedisjointpartitioningofthedatawhereaseachdatapointhasacertainprobabilityofbelongingtoeachofthepartitionsi
此文档下载收益归作者所有