維吾爾語語音識別技術在智能家居中的研究與應用.pdf

1、碩士研究生學位論文新疆大學論文題目論文題目(中文中文)：維吾爾語語音識別技術在智能家居中的：維吾爾語語音識別技術在智能家居中的研究與應用研究與應用論文題目論文題目(外文外文)：ResearchesApplicationsofUyghurSpeechRecognitionTechnologyintheSmartHome研究生姓研究生姓名：吾提庫爾：吾提庫爾艾爾肯艾爾肯學科、專學科、專業(yè)：計算機應用技術：計算機應用技術研究方向：語音信號處理

2、：語音信號處理導師姓名職導師姓名職稱：吾守爾：吾守爾斯拉木斯拉木院士、教授院士、教授論文答辯日期2017年5月日學位授予日期2017年6月日IIAbstractTraditionalappliancestraditionalhouseholdequipmenthavebecomeoutdated.Soifthereisawaytomaketheoldappliancesdonotchangeinthecaseoftheuseofsmar

3、tphonestoachieveUighurspeechcontrolcontrolhowgoodah!InthisideaIstartedtheresearchwkinthisroundtheintelligenthomesystemisbasedonthefamilyusingpatternrecognitioncontrolinstructionmappingcommunicationwkfmationspeechrecognit

4、ionsemanticunderstingthirdpartyintelligentcontrolcomponents.AtpresentinthemarketthereisnoUighurvoicecontrolofintelligenthomeproducts.UighurspeechrecognitiontechnologyhasmaturedinrecentyearsXinjiangUniversityledbyagroupof

5、researchersinthelongtermefftsofXinjianguniversitytoaccumulatetherelevantfieldsofkeytechnologies.InthefieldofUighurspeechrecognitionresearchtheCollegeofInfmationScienceengineeringofXinjiangUniversityhasstrongscientificres

6、earchabilityhasaccumulatedovertheyears.Iwasinvolvedinthelabatyspeechrecognitionrelatedprojectsinthegraduatestagebenefitedalotfromit.Iusethreeyearsofstudytimetointegratetheknowledgeofspeechrecognitionintothebackgroundofin

7、telligenthomehardwareequipmenttheUighurspeechrecognitiontechnologyappliedtothefieldofintelligenthome.ThroughthespeechtechnologytorecognizetheintegrationoftechnologyintelligenthometechnologyintheoldhomeappliancesUighurspe

8、echcontrolsetupaconvenientintelligenthomeControlclientapplicationmakingthesystemmeintelligenthumane.ThispaperfocusesonthespeechrecognitionbasedontheCMUSphinx4.0UygurLanguagetherealizationofthevoicecontrolelectricalequipm

9、entthroughtheIntelligentHomeHardwareinterface(TVlightbulbhotwaterkettlewaterheateretc.throughtheswitchcontrolofelectricalequipmentthroughremotecontrolofelectricalequipment).ThespeechrecognitionprocessofUighurphraseconsis

10、tsoffourparts(speechsignalfeatureparameterextractionbuildingacousticmodelconstructinglanguagemodeldecoderpartpatternmatching).ThispapermakesadetailedcomparisonanalysisofthefourrelatedtechnologieschoosesMFCChmmNGramtechni