Abstract. Spectral subband centroids (SSC) have been used as an additional feature to cepstral coefficients in speech and speaker recognition. SSCs are computed as the centroid frequencies of subbands and they capture the dominant frequencies of the short-
SpeakerVeri cationwithAdaptiveSpectral
SubbandCentroids
TomiKinnunen1,BingjunZhang2,JiaZhu2,andYeWang2
SpeechandDialogueProcessingLab
InstitutionforInfocommResearch(I2R)
21HengMuiKengTerrace,Singapore119613
ktomi@i2r.a-star.edu.sg
2DepartmentofComputerScience
SchoolofComputing,NationalUniversityofSingapore(NUS)
3ScienceDrive2,Singapore117543
{bingjun,zhujia,wangye}@comp.nus.edu.sg1
Abstract.Spectralsubbandcentroids(SSC)havebeenusedasanaddi-tionalfeaturetocepstralcoe cientsinspeechandspeakerrecognition.SSCsarecomputedasthecentroidfrequenciesofsubbandsandtheycapturethedominantfrequenciesoftheshort-termspectrum.InthebaselineSSCmethod,thesubband ltersarepre-speci ed.Toallowbet-teradaptationtoformantmovementsandotherdynamicphenomena,weproposetoadaptthesubband lterboundariesonaframe-by-framebasisusingagloballyoptimalscalarquantizationscheme.Themethodhasonlyonecontrolparameter,thenumberofsubbands.Speakerveri- cationresultsontheNIST2001taskindicatethattheselectionoftheparameterisnotcriticalandthatthemethoddoesnotrequireadditionalfeaturenormalization.
1Introduction
Theso-calledmel-frequencycepstralcoe cients[1](MFCC)haveproventobee cientfeaturesetforspeakerrecognition.Aknownproblemofcepstralfeatures,however,isnoisesensitivity.Forinstance,convolutivenoiseshiftsthemeanvalueofthecepstraldistributionwhereasadditivenoisetendstomodifythevariances
[2].Tocompensateforthefeaturemismatchbetweentrainingandveri cationutterances,normalizationsinfeature,modelandscoredomainsarecommonlyused[3].
Spectralsubbandcentroids[4;5;6;7](SSC)areanalternativetocepstralcoe cients.SSCsarecomputedasthecentroidfrequenciesofsubbandspectraandtheygivethelocationsofthelocalmaximaofthepowerspectrum.SSCshavebeenusedforspeechrecognition[4;5],speakerrecognition[7]andaudio ngerprinting[6].RecognitionaccuracyofSSCsislowerinnoise-freecondi-tionscomparedwithMFCCs.However,SSCscanoutperformMFCCsinnoisyconditionsandtheycanbecombinedwithMFCCstoprovidecomplementaryinformation.
SPEAKER VERIFICATION WITH LOW STORAGE REQUIREMENTS Center for Spoken Language ...The most common approach to text-dependent speaker veri cation is spectral-...
Speaker Verification wit... 暂无评价 8页 免费 Adaptive spectral method...... MA 01803-3304 ABSTRACT Dispersive transform spectral imagers with both one-...
Semi-Blind Spectral Deconvolution with Adaptive Tikhonov Regularization_化学_自然科学_专业资料。Semi-Blind Spectral Deconvolution with Adaptive Tikhonov ...
2014 Adaptive total variation-based spectral deconvolution with the split Bregman method_计算机软件及应用_IT/计算机_专业资料。Adaptive total variation-based ...
Adaptive Techniques For ... 暂无评价 63页 免费 Laplacian eigenmaps and ... 7页 免费 Spectral Analysis with S... 暂无评价 6页 免费 C--ALGEBRAS AND...
Adaptive spectral methods for simulation output analysis This paper addresses two central problems in simulation methodology: the generation of conjidence ...
speakerverificationinnoiseconditions.AstheMelspectralresolutionin frequencyhas a logarithmicrelationshipwithlinearcan frequency,avoidhar- high frequencydomainwoulddecline...
DIGITAL SPECTRAL ANALYSIS MATLAB for Adaptive Filtering and Prediction LECTURE ... microphone close to the noise source and reasonably far from the speaker....
Speaker Verification Based on G.729 and G.723...Entries from the ?xed, adaptive, and gain ...with those of spectral envelopes for speaker veri...
the adaptive MV-SVD outperforms the spectral subtraction at low SNR ...Applying NSS with parameter optimization for each speaker 49.7% WER were ...
热门文档