预览加载中，请您耐心等待几秒...

基于支持向量数据描述的说话人识别研究的综述报告.docx

预览
1/2
2/2

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

基于支持向量数据描述的说话人识别研究的综述报告 Abstract Speakerrecognitionhasemergedasavitalandchallengingtaskinthefieldofsignalprocessingandpatternrecognitionduetoitsextensivepracticalapplicationsintheauthenticationandidentificationprocess.Inthispaper,acomprehensivereviewoftherecentliteratureonspeakerrecognitionispresented,especiallyfocusingonthesupportvectordatadescription-basedapproaches.Thepaperdefinesthespeakerrecognitionmethodologies,evaluatestheirperformance,andhighlightstheadvancesinthesupportvectordatadescription-basedapproach. Introduction Speakerrecognitionisabiometricverificationprocessthatidentifiesandverifiesindividualsbasedontheirspeechcharacteristics.Thespeakerrecognitiontechniquescanbedividedintotwocategories:text-dependentandtext-independent.Text-dependenttechniquesrequirethespeakerstouseaspecificfixedtextpattern,whiletext-independenttechniquesusefreelyspokenutterances.Thetext-independenttechniquesaremoresuitableforreal-worldapplicationsbecausetheyoffermoreflexibilityandaccuracy. SupportVectorDataDescription SupportVectorDataDescription(SVDD)isaclassicmethodologyproposedbyTaxandDuinin1999,whichhasbeenextensivelyusedforspeakerrecognitioninrecentyears.SVDDisaone-classclassificationalgorithmthataimstodefineaboundaryaroundthetargetclasswhileminimizingthesupportvectors.TheSVDDalgorithmisdesignedtofindthesmallesthyperspherethatcanenclosethedatavectorsbelongingtoaspecificclassandexcludeallotherdatavectorsfromthehypersphere. SVDD-BasedSpeakerRecognitionApproaches SeveralspeakerrecognitionapproachesbasedonSVDDhavebeenproposedinrecentyears.Theseapproachesdifferinthefeatureextractiontechnique,datapreprocessing,andclassificationschemeused.Oneofthewell-knownmethodsistheKernel-PCA-SVDD(KPCA-SVDD)method,whichuseskernelprincipalcomponentanalysistoextractthefeatures,followedbytheSVDDalgorithmtoclassifytheextractedfeatures.AnotherapproachisusingRadialBasisFunction(RBF)networksandC-SVDDinatwo-stagemanner,whichusesRBFtoextractthefeaturesandSVDDtoclassifythefeatures.AnensembleofSVDDswithdiff