预览加载中,请您耐心等待几秒...
1/2
2/2

在线预览结束,喜欢就下载吧,查找使用更方便

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

改进的基于MELP的非连续传输语音编码算法 Abstract: Inrecenttimes,theincreasingimportanceofvoicecommunicationhasdriventhedevelopmentofimprovedspeechcodingalgorithms.Manyspeechcodingapproacheshavebeenproposed,focusingondifferentdegreesofcompressionwhilemaintainingacceptablespeechquality.AmongthesealgorithmsistheMELPspeechcodingalgorithm,whichprovideslowbit-ratecompressionwhileofferinggoodspeechquality.Thispaperproposesanewnon-continuoustransmissionspeechcodingalgorithmbasedontheMELPcodingapproach.TheproposedalgorithmincludesseveralimprovementstotheexistingMELPalgorithm,includingnewsignaltransformtechniquesandspectralsubtractionprocedures. Introduction: Speechcoding,orspeechcompression,isacrucialtaskinmoderncommunicationsystems,wherethetransmissionofspeechsignalsisincreasinglyprioritized.Speechcodingtechniquesaimtoreducethebit-rateofspeechsignalswhileensuringthattheoutputsignalmaintainsanacceptablelevelofspeechquality.Awiderangeofspeechcodingalgorithms,withdifferentdegreesofcompressionandspeechquality,havebeenproposedintheliterature.Amongthem,MELPcodingisanattractiveapproach,offeringalowbit-ratecompressionwhileretaininganacceptablelevelofspeechquality. However,theMELPalgorithmstillhassomelimitationsthatlimititsperformanceincertainapplications.Inthispaper,weproposeanewnon-continuoustransmissionspeechcodingalgorithm,basedontheMELPcodingapproach,thatovercomessomeoftheMELPalgorithm'slimitations.Inthenewalgorithm,weintroduceseveralsignaltransformationtechniquesandspectralsubtractionprocedurestoimprovethespeechsignal'squality. MELPAlgorithmvs.ProposedAlgorithm: TheMELPalgorithmisalowbit-ratespeechcodingalgorithmthatusesLinearPredictionCoding(LPC)andCodeExcitedLinearPrediction(CELP)algorithms.Italsousesvectorquantizationtoreducethebit-rate.AlthoughMELPprovidesacceptablespeechquality,itstillhassomedrawbacksinsomeapplications,suchaslow-qualitysignalreconstructionduetothelimitationofthecoderincapturingsignalvariations. Inourproposedalgorithm,weimprovethesignaltransformationstagebyincludingaFastFou