Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/17884
DC FieldValueLanguage
dc.contributor郭仁泰zh_TW
dc.contributor邱國欽zh_TW
dc.contributor.advisor李宗寶zh_TW
dc.contributor.author林凱昇zh_TW
dc.contributor.authorLin, Kai-Shengen_US
dc.contributor.other中興大學zh_TW
dc.date2008zh_TW
dc.date.accessioned2014-06-06T07:02:23Z-
dc.date.available2014-06-06T07:02:23Z-
dc.identifierU0005-2606200712432100zh_TW
dc.identifier.citation[1]M. Bilginer Gulmezoglu, Vakif Dzhafarov, and Atalay Barkana ,“The common vector approach and its relation to principal component analysis” IEEE Trans. On Speech and Audio Processing, vol. 9. No. 6, 1999 [2]M. Keskin, M. B. Gulmezoglu, O. Parlaktuna, and A. Barkana, “Isolated word recognition by extracting personal differences,” in Proc. 6 th Int.Conf. Signal Processing Applications and Technology, Boston, MA , pp.1989-1992, 1996. [3]S. Yucel, “Application of Gram-Schmidt orthogonalization method to speech recognition for different noise levels” graduation project, Elect. Electron. Eng. Dept., Osmangazi Univ., Eskisehir, Turkey, 1996. [4]H. Angm, “Common vector obtained from linearly independent speech vectors by using LPC parameters,” graduation project, Elect. Electron. Eng. Dept., Osmangazi Univ., Eskisehir, Turkey, 1995. [5]Hakan Cevikalp and Mitch Wilkes,”Discriminative Common Vectors for Face Recognition”pattern analysis and machine intelligence,vol.27,NO.1, 2005 [6]L. Rabiner and B. H. Juang,”Fundamentals of Speech Recognition”, Englewood Cliffs,NJ:prentice-Hall,1993. [7]Frederick Jelinek ,”Statistical methods for speech recognition ”, Mass.:/MIT Press,Cambridge,1998 [8]李宗寶,吳宗憲。” 探討Kmean之共同向量法應用於國語數字辨識”。碩士論文,國立中興大學應用數學研究所,台中,2005。 [9]王小川。”語音訊號處理”。台北市:全華,2004。 [10]李宗寶,林靖剛。”利用Multiple Common Vector 於國語數字之語音辨識”。 碩士論文,國立中興大學應用數學研究所,台中,2006。zh_TW
dc.identifier.urihttp://hdl.handle.net/11455/17884-
dc.description.abstract摘要 本篇論文主要是利用「Discriminative Common Vectors法」,來探討國語數字0 ~ 9的語音辨識,「Discriminative Common Vectors法」原先是使用在臉部辨識的方法之一,在辨識率上有一定的水準,所以本人試著套用此方法在語音辨識上,利用臉部辨識的方法來做語音的辨識,找出最相似的語音,看效果是否會較好。 本實驗中的語音資料庫,由21人(12名男生,9名女生)所建構,每人唸0 ~ 9各三次,前兩次做為訓練語音模型用,第三次則做為特定語者辨識時測試用,接著對所有語音資料做前處理,並建構模型來做為辨識比對。 本論文所要討論影響辨識率的因子有音框取樣個數、子母音的比例分配兩種。在特定語者的辨識結果最高為取樣音框個數為200,子母音分配比例為1.5:8.5的狀況下,辨識率最高為92.86%。zh_TW
dc.description.tableofcontents摘要 i 附圖目錄 v 第一章 緒論..........................................1 1.1研究動機...................................1 1.2國內外相關研究.............................2 1.3語音辨識介紹...............................4 1.4語音辨識方法概述...........................6 1.4.1語音前處理與特徵值的求取..............6 1.4.2音框數壓縮與擴展......................9 1.4.3訓練出語音模型和辨識比對..............10 1.5論文架構...................................12 第二章 語音訊號的前處理與特徵值......................13 2.1前言.......................................13 2.1.1子音與母音............................13 2.1.2 基頻.................................14 2.2語音的前處理...............................15 2.2.1數位化取樣............................15 2.2.2常態化................................17 2.2.3語音端點偵測..........................19 2.2.4切割音框..............................21 2.2.5預強調................................22 2.2.6視窗化................................22 2.3特徵值的求取...............................24 2.3.1自相關函數............................25 2.3.2線性預估編碼係數(LPC)求法...........25 2.3.3倒頻譜係數............................27 第三章 語音模型的建立與辨識方法......................29 3.1前言.......................................29 3.2音框的壓縮與擴展...........................29 3.2.1音框數超過Y...........................31 3.2.2音框數少於Y...........................31 3.2.3音框數等於Y...........................32 3.3建構模型的方法.............................32 3.3.1共同向量法原理........................32 3.3.2使用 的Range Space來獲得Discriminative Common Vectors........................36 3.3.3使用差異子空間與Gram-Schmidt Orthogonalization Procedure來獲得 Discriminative Common Vectors.........41 3.3.4結論..................................42 3.4辨識的方法.................................42 3.4.1待測語音的處理.........................42 3.4.2比對的方法............................43 第四章 實驗操作流程與實驗結果........................45 4.1操作介面...................................45 4.2實驗流程...................................45 4.2.1語音來源..............................45 4.2.2影響辨識率的可能因子..................46 4.2.3辨識結果..............................46 第五章 結論與未來展望................................65 5.1結論.......................................65 5.2未來展望...................................65 參考文獻.............................................67 圖1.1 語音流程圖.....................................4 圖1.2倒頻譜參數流程圖................................8 圖1.3 語音前處理及求取特徵值.........................9 圖1.4 壓縮與擴展.....................................10 圖1.5 辨識流程圖.....................................11 圖2.1子音和母音的聲波圖..............................14 圖2.2語音類比訊號圖..................................16 圖2.3數位取樣訊號圖..................................16 圖2.4原始語音「6」的波形圖...........................18 圖2.5經過常態化後,語音「6」的波形圖.................18 圖2.6偵測語音端點流程圖..............................20 圖2.7取樣音框有部分重疊..............................22 圖2.8語音取漢明窗....................................24 圖3.1音框壓縮與擴展..................................30 圖3.2三維特徵空間的共同向量 .........................36 圖4.1總音框數30,子母音比例(2.25:7.75)的辨識率....47 圖4.2總音框數50,子母音比例(2.25:7.75)的辨識率....47 圖4.3總音框數70,子母音比例(2.25:7.75)的辨識率....48 圖4.4總音框數90,子母音比例(2.25:7.75)的辨識率....48 圖4.5總音框數110,子母音比例(2.25:7.75)的辨識率...49 圖4.6總音框數200,子母音比例(2.25:7.75)的辨識率...49 圖4.7總音框數30,子母音比例(2:8)的辨識率..........50 圖4.8總音框數50,子母音比例(2:8)的辨識率..........50 圖4.9總音框數70,子母音比例(2:8)的辨識率..........51 圖4.10總音框數90,子母音比例(2:8)的辨識率.........51 圖4.11總音框數110,子母音比例(2:8)的辨識率........52 圖4.12總音框數200,子母音比例(2:8)的辨識率........52 圖4.13總音框數30,子母音比例(1.75:8.25)的辨識率...53 圖4.14總音框數50,子母音比例(1.75:8.25)的辨識率...53 圖4.15總音框數70,子母音比例(1.75:8.25)的辨識率...54 圖4.16總音框數90,子母音比例(1.75:8.25)的辨識率...54 圖4.17總音框數110,子母音比例(1.75:8.25)的辨識率..55 圖4.18總音框數200,子母音比例(1.75:8.25)的辨識率..55 圖4.19總音框數30,子母音比例(1.5:8.5)的辨識率.....56 圖4.20總音框數50,子母音比例(1.5:8.5)的辨識率.....56 圖4.21總音框數70,子母音比例(1.5:8.5)的辨識率.....57 圖4.22總音框數90,子母音比例(1.5:8.5)的辨識率.....57 圖4.23總音框數110,子母音比例(1.5:8.5)的辨識率....58 圖4.24總音框數200,子母音比例(1.5:8.5)的辨識率....58 圖4.25總音框數30,子母音比例(1.25:8.75)的辨識率...59 圖4.26總音框數50,子母音比例(1.25:8.75)的辨識率...59 圖4.27總音框數70,子母音比例(1.25:8.75)的辨識率...60 圖4.28總音框數90,子母音比例(1.25:8.75)的辨識率...60 圖4.29總音框數110,子母音比例(1.25:8.75)的辨識率..61 圖4.30總音框數200,子母音比例(1.25:8.75)的辨識率..61 表4.1語音辨識率......................................62 圖4.31子母音比例(2.25:7.75)的辨識率折線圖.........63 圖4.32子母音比例(2:8)的辨識率折線圖...............63 圖4.33子母音比例(1.75:8.25)的辨識率折線圖.........63 圖4.34子母音比例(1.5:8.5)的辨識率折線圖...........64 圖4.35子母音比例(1.25:8.75)的辨識率折線圖.........64zh_TW
dc.language.isoen_USzh_TW
dc.publisher應用數學系所zh_TW
dc.relation.urihttp://www.airitilibrary.com/Publication/alDetailedMesh1?DocID=U0005-2606200712432100en_US
dc.subjectDiscriminative Common Vectoren_US
dc.subject共同向量zh_TW
dc.subject特徵值zh_TW
dc.title利用 Discriminative Common Vectors 於國語數字之語音辨識zh_TW
dc.titleUsing the Method of Discriminative Common Vectors to Speech Recognition of Mandarin digitsen_US
dc.typeThesis and Dissertationzh_TW
item.cerifentitytypePublications-
item.openairecristypehttp://purl.org/coar/resource_type/c_18cf-
item.languageiso639-1en_US-
item.fulltextno fulltext-
item.grantfulltextnone-
item.openairetypeThesis and Dissertation-
Appears in Collections:應用數學系所
Show simple item record
 

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.