跳到主要內容

臺灣博碩士論文加值系統

(54.91.62.236) 您好!臺灣時間:2022/01/18 00:21
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:廖家慶
研究生(外文):Chia-Ching Liau
論文名稱:語者調適之應用研究
論文名稱(外文):The Research of Speaker Adaptation
指導教授:莊堯棠
指導教授(外文):Yau-Tarng Juang
學位類別:碩士
校院名稱:國立中央大學
系所名稱:電機工程研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2002
畢業學年度:90
語文別:中文
論文頁數:58
中文關鍵詞:語者調適
外文關鍵詞:speaker adaptation
相關次數:
  • 被引用被引用:1
  • 點閱點閱:125
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
摘 要
在語音辨識系統中,特定語者(Speaker-dependent)型語音辨識系統雖有高辨識率的優點,但當應用到新語者時須花釵h語音訓練資料和時間;而不限語者(Speaker-independent)或多語者(multi-speaker)型的語音辨識系統,除最初建立系統時所需語音資料外,應用於新語者時不再需新語音訓練資料,但其辨識率普遍不高。語者調適(Speaker-adaptive)辨識系統則利用一充分訓練過的參考系統已知資訊,藉新語者少量語音資料訓練 ,可達到接近特定語者系統的辨識率,因此論文中將針對語者調適系統進行研究。
本論文內容包含兩個主要研究主軸,其一為如何在少量調適語料之狀況下,增進改善調適演算法,藉此提升系統辨識率與調適結果;另一主軸則為利用增進後之調適演算法實際應用於線上辨識與調適。
於第一研究主軸中,其重點在於考慮初始模型與最大可能性線性迴歸(Maximum Likelihood Linear Regression,MLLR)兩者間貢獻的比重分配,藉由找出最佳平衡點來提升調適性能。接著並考慮向量場平滑化(Vector-Field-Smoothing,VFS)轉移向量場的調適方式,針對沒有觀測到之調適語料模型,加以參考有調適語料之模型來進行調整,藉此特性再搭配權重化之MLLR調適方法研究其調適效果。接者利用特定語者模型與不特定語者模型來架構出特徵向量空間,由此特徵向量空間來找出語者的代表點所在,藉此調整系統模型參數。而在第二研究主軸內,藉由所發展出少量調適語料即能達到調適系統之演算法,將此調適演算法應用於線上系統,使語者能夠感受到辨識與調適之即時變化。
Speaker adaptation has been applied to speech recognition to get a speaker dependent system with a good performance. Most adaptation techniques use the initial model as a starting point and then introduce speaker’s specific information. By using the adapted parameters, the recognition performance can be significantly improved.
In this thesis, we present a variation on improving the performance of maximum likelihood linear regression (MLLR) in cases of little adaptation data. The transformed Gaussian means are interpolated with the means in the initial mean models. The VFS algorithm proposed by the following steps. First, the transfer vectors are estimated. Then, interpolation and smoothing are performed using the transfer vectors. We applied the idea of using eigenvoices, a set of orthogonal basis vectors derived from the parameters of speaker dapendent models trained on reference speakers.
目 錄
摘 要…………………………………………………………i
附圖目錄……………………………………………………iv
附表目錄………………………………………………………v
第一章 序論…………………………………………………1
1.1語音處理發展與應用…………………………………………1
1.2語者調適簡述…………………………………………………2
1.3研究動機………………………………………………………3
1.4研究目標………………………………………………………4
1.5論文大綱………………………………………………………5
第二章 語音處理基本技術…………………………………6
2.1特徵參數求取…………………………………………………6
2.2隱藏式馬可夫模型……………………………………………9
第三章 語者調適相關技術…………………………………16
3.1修正最大可能性線性迴歸…………………………………16
3.1.1最大可能性線性迴歸理論……………………………16
3.1.2 權重化MLLR調適方法之推導………………………17
3.1.3最大可能性線性迴歸其對角化 之推導……………20
3.2向量場平滑化(VFS) ………………………………………24
3.3權重化MLLR調適方法與向量場平滑化之合併應用………26
3.4特徵語音調適法(Eigenvoice) ……………………………27
第四章 系統模型架構………………………………………29
4.1系統次音節模型架構…………………………………29
4.2系統次音節模型訓練與辨識………………………………32
第五章 研究方法與結果…………………………………………36
5.1研究環境………………………………………………………36
5.1.1系統特徵參數與研究設備……………………………36
5.1.2訓練、調適及測試語料…………………………………36
5.2權重化修正MLLR調適之研究………………………………37
5.3權重化修正MLLR與VFS結合調適之研究………………40
5.4特徵語音調適法之研究………………………………………42
5.5辨識與調適應用系統…………………………………………44
第六章 結論與展望…………………………………………47
6.1結論……………………………………………………………47
6.2展望……………………………………………………………48
參考文獻……………………………………………………49
參考文獻[1] Rabiner,L. R. et al.”Recognition of Isolated Digits Using HiddenMarkov Models with Continuous Mixture Densities,” AT&T Technical Journal 64(6):1211-1233,1985.[2] Juang,B.H.,and Rabiner,L.R.”Mixture Auto-regressive Hidden Markov Models for Speech Signals,”IEEE Trans.on ASSP,vol.33,No.6,pp.1404-1413,Dec.1985.[3] Rabiner,L.R.,and Juang,B.H. ”An Introduction to Hidden Markov Models”IEEE ASSP Magzine,Jan.1986.[4] L.R. Rabiner, “A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition”. Proc. IEEE, Vol. 77, No.2, pp. 257—286, Feb. 1989.[5] C-H. Lee, C-H. Lin, and B-H. Juang, “A Study on Speaker Adaptation of the Parameters of Continuous Density Hidden Markov Models”. IEEE Trans. on Sig. Proc., Vol. 39, No. 4, pp. 806—814, April 1991.[6] Heidi Christensen, “Speaker Adaptation of Hidden Markov Models using Maximum Likelihood Linear Regression”. MSc.E.E. Thesis. Aalborg University, Denmark, June 1996.[7] C.J. Leggetter and P.C. Woodland, “Speaker Adaptation of HMM’s using Linear Regression”. Technical Report GUED/F-INFENG/ TR.181, Cambridge University, June 1994.[8] C.J. Leggetter and P.C. Woodland, “Maximum Likelihood Linear Regression for Speaker Adaptation of Continuous Density Hidden Markov Models”. Computer Speech and Language, Vol. 9, pp. 171—185, 1995.[9] C.J. Leggetter and P.C. Woodland, “Flexible Speaker Adaptation using Maximum Likelihood Linear Regression”. Proc. ARPA Spoken Language Technology Workshop, pp. 104—109, Feb. 1995.[10] C.J. Leggetter and P.C. Woodland, “Speaker Adaptation of continuous density HMMs using Multivariate Linear Regression”. ICSLP-94, Vol. 2, pp. 451—454, Yokohama, 1994.[11] M.J.F. Gales, “Maximum Likelihood Linear Transformation for HMM-Based Speech Recognition”. Technical Report GUED/F-INFENG/TR.291, Cambridge University, May 1997.[12] M.J.F. Gales, “The Generation and use of Regression Class Trees for MLLR Adaptation”. Technical Report GUED/F-INFENG/TR.263, Cambridge University, August 1996.[13] A. Sankar and C-H. Lee, “A Maximum-Likelihood Approach to Stochastic Matching for Robust Speech Recognition”. IEEE Trans. on Speech and Audio Proc., Vol. 4, pp. 190—202, May 1996[14] L. R. Rabiner and R. W. Schafer, “ Digital Processing of Speech Recognition Signals ”, Prentice-Hall Co. Ltd, 1978.[15] M. Tonomura, T. Kosaka and S. Matsunaga, “Speaker Adaptation Based on Transfer Vector Field Smoothing using Maximum a Posteriori Probability Estimation”. ICASSP-95, Vol. 1, pp. 688—691, 1995.[16] B.F. Necioglu, M. Ostendorf, and J.R. Rohlicek, “A Bayesian Approach to Speaker Adaptation for the Stochastic Segment Model”. ICASSP-92, Vol. 1, pp. 437—440, 1992.[17] J-I. Takahashi and S. Sagayama, “Fast Telephone Channel Adaptation Based on Vector Field Smoothing Technique”. Second IEEE Workshop on Interactive Voice Technology for Telecommunications Applications, pp. 97—100, 1994.[18] J. Takahashi and S. Sagayama, “Vector-Field-Smoothed Bayesian Learning for Incremental Speaker Adaptation”. ICASSP-95, Vol. 1, pp. 696—699, 1995.[19] J. Takahashi and S. Sagayama, “Minimum Classification Error Training for a Small Amount of Data Enhanced by Vector-Field-Smoothed Bayesian Learning”. ICASSP-96, Vol.: 2, pp. 597—600, 1996.[20] R. Kuhn, P. Nguyen, J. —C. Junqua, N. Niedzielski, “Rapid Speaker Adaptation in Eigenvoice Space”. IEEE Trans. on Speech and Audio Proc., Vol. 8, pp. 695-707, Nov. 2000.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top