(18.206.187.91) 您好!臺灣時間:2021/05/19 02:25
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:李佳慧
研究生(外文):Jia-Hui Li
論文名稱:不特定語者國語語音字詞辨識系統研究
論文名稱(外文):Research of Speaker Independent Spoken MandarinSpoken Word Recognition System
指導教授:杜筑奎
指導教授(外文):Chu-Kuei Tu
學位類別:碩士
校院名稱:中原大學
系所名稱:資訊工程研究所
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:2001
畢業學年度:89
語文別:中文
論文頁數:63
中文關鍵詞:不特定語者隱藏式馬可夫模型向量量化
外文關鍵詞:HMMvector quantizationspeaker-independent
相關次數:
  • 被引用被引用:5
  • 點閱點閱:91
  • 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:2
本篇論文在建立一套能辨識台灣鐵路局站名之不特定語者語音辨識系統,此系統在Pentium PC、Windows 98作業系統下,利用Microsoft Visual C++ 6.0與Intel Recognition Primitives Library 等工具開發。
在聲學模型訓練過程,系統利用能量凹處(energy dip)、越零率(zero crossing rate)、自相關係數(autocorrelation function)做語音切割後,使用梅爾倒頻譜係數(MFCC, mel scale filter cepstral coefficient)求取特徵參數,經過Binary splitting尋找向量量化碼本,並以離散型隱藏式馬可夫模型(DHMM, Discrete Hidden Markov Models)建立語音模型後,再使用波氏演算法(BaumWelch Algorithm)做調適。在辨認方面,則是採用Kohonon Network求取碼字序列(codeword sequence),以光束搜尋法(Beam Search)取代維特比演算法(Viterbi algorithm)來計算最佳辨認的機率。
在不特定語者的實驗中,辨識正確率可到達85.75%,證明是一套可行的研究方法。

In this thesis, a speaker-independent Mandarin spoken word recog-nition system for Taiwan railway station is implemented. The system is built with components that include a Pentium PC, Microsoft Windows 98 operation system, Microsoft Visual C++ 6.0 and Intel Recognition Primi-tives Library.
During the acoustic-model training stage, we employ the energy dip, zero crossing rate, and autocorrelation function to segment speech sounds. And use the MFCC (mel scale filter cepstral coefficient) to evaluate fea-ture parameters. Through the process of Binary splitting the vector quan-tization codebooks are found, the DHMM (Discrete Hidden Markov Models) is used to establish all acoustic-models, and the BaumWelch al-gorithm is chosen to adapt the optimal solution. On the recognition part, the Kohonon Network is used to calculate codeword sequence. The Beam search is used to replacement of Viterbi algorithm that gives the best re-sult of recognition in DHMM.
The recognition rates of speaker-independent experiments can reach up to 85.75%. It shows that the system has achieved good performance.

目錄
第一章緒論
1.1研究動機 1
1.2國語語音的性質 2
1.3研究目標 5
1.4章節大要 6
第二章理論基礎
2.1 語音辨識架構與類型 7
2.2 能量函數/越零率/一階反射係數 11
2.3 梅爾倒頻譜參數 14
2.4 向量量化 19
2.5 隱藏式馬可夫模型 23
2.6 光束搜尋演算法 26
2.7 波式演算法 31
第三章語音辨識系統建立
3.1 語音辨識系統架構 32
3.2 特徵參數抽取 34
3.3 特徵參數向量量化 36
3.4 語音模型建立與訓練 38
3.5 語音辨識與調適 40
第四章實驗結果與比較
4.1 實驗架構 43
4.2 特定語者模型之實驗比較 45
4.3 不特定語者模型之實驗比較 50
第五章結論 55
附錄 58
參考文獻 61

[1] Lin-shan Lee, Yumin Lee, "Voice Access of Global Information for Broadband Wireless: Technologies of Today and Challenges of Tomorrow" , on Proceedings of the IEEE, Vol. 89 Issue.1 , pp. 41 -57 ,January 2001 [2] Jhing-Fa Wang; Jia-Ching Wang; Jar-Ferr Yang; Jian-Jia Wang ,”A voicing- driven packet loss recovery algorithm for analysis-by-synthesis predictive speech coders over internet” , Multimedia, IEEE Transactions on , Vol. 3 Iss. 1 , pp. 98 —107, March 2001[3] Yuan-Fu Liao; Sin-Horng Chen , “A modular RNN-based method for continuous Mandarin speech recognition”, Speech and Audio Processing, IEEE Transactions on , Vol. 9 Iss. 3 , pp. 252 —263, March 2001 [4] Chen, S.-H.; Wang, J.-F. , “Application of wavelet transforms for C/V segmenta- tion on mandarin speech signals”, Vision, Image and Signal Processing, IEEE Proceedings , Vol. 148 Iss. 2 , pp.133 —139, Apr 2001 [5] T. H. Hwang and H. C. Wang, “A fast algorithm on parallel model combination for noisy speech recognition,” Computer Speech and Language, vol. 14, pp. 81-100, 2000. [6] Chi-Min Liu; Chin-Chih Chiu; Hung-Yuan Chang ,“Design of vocabulary indepe- ndent Mandarin keyword spotters”, Speech and Audio Processing, IEEE Transactions on , Vol. 8 Iss. 4 , pp. 483 —487, July 2000 [7] 中華民國計算語言學學會, “2001語音訊號研討會會刊”, Taiwan, April 2001[8] Intel Co.,Intel Recognition Primitives Library Reference Manual,2000[9] 謝國平,語音學概論 (增訂版),1998,三民書局[10] 羅肇錦,國語學,1980,五南書局[11] 許隆倫,基於右相聯音素模型的中文連續語音辨識之語者問題研究,1995,台灣大學 電機工程學研究所碩士論文[12] 謝宏昀,以單一階段整合聲學、語言學和音韻學的極大字彙國語連續語音辨識架 構”,1996,台灣大學資訊工程學研究所碩士論文[13] 陳佳欣,藉助界標知識及調整動態特徵參數之國語音節辨識,1997,台灣大學電機工 程學研究所碩士論文[14] 林宸生,數位信號-影像與語音處理,1996,全華書局[15] J.R. Deller, “Discrete-Time Processing of Speech Signals”, Macmillan, 1993[16] L.R. Rabiner and B.H. Juang,”Fundamentals of Speech Recognition”, Prentice Hall,1993[17] Shigeki Sagayama, Satoshi Takahashi, "On The Use of Scalar Quantization for Fast HMM Computation", ICASSP, Vol. 1, pp. 213-216, 9-12 May 1995.[18] T. Kohonen, G. Barna, and R. Chrisley,,”Statistical pattern recognition with neural networks: Benchmarking studies.”,IEEE Int. Conf. On Neural Networks,vol. I, pp.61-68, 1988[19] T.Kohonen, “Self-Organization and Associative Memory(3nd Ed.)”,Springer- Verlag, New York, Berlin, Heidelberg, 1989[20] X. D. Huang, Y. Ariri, and M. A. Jack. “Hidden Markov Models for Speech Recognition.”, Number 7 in 1nformtion Technology Series. Edinburgh University Press, Edinburgh, 1990.[21] J.-T. Chien, “Online hierarchical transformation of hidden Markov models for speech recognition”, IEEE Transactions on Speech and Audio Processing, vol. 7, no. 6, pp. 656-667, November 1999.[22] R. Haeb-Umbach.; H. Ney, "Improvements in Beam Search for 10000-Word Continuous-Speech Recognition," IEEE Trans. On Speech and Audio Processing,Vol. 2,pp. 353-356,April 1994.[23] Xiaolin Li, Marc Parizeau and Rëjean Plamondon, “Training Hidden Markov Models with Multiple Observations--A Combinatorial Method”,IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 22, No. 4, April 2000 [24] 王新民,極大字彙國語連續語音辨識:金聲三號(工作站版),1994,台灣大學電機工程 學研究所碩士論文[25] 王漢蘭,隱藏式馬可夫模型應用於語音訊號辨識之研究,1999,中原大學資訊工程學 研究所碩士論文[26] 楊永泰,隱藏式馬可夫模型應用於中文語音辨識之研究,2000,中原大學資訊工程學 研究所碩士論文

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top