Please use this identifier to cite or link to this item:
標題: 利用權重式多重KNN法於中字彙之特定語者中文單音辨識
Using the Method of Weighted multiple-KNN to Recognize Isolated Mandarin Word for Speaker-Dependent System
作者: 籃元隆
Lan, Yuan-Lung
關鍵字: K-Nearest Neighbor;K-最近鄰居法
出版社: 應用數學系所
引用: [1]王小川(2004)。 “語音訊號處理”。台北市:全華。 [2]王國榮(2000)。 “Visual Basic 6.0 實戰講座”。台北市:旗標。 [3]吳明哲,黃世陽(1998)。 “Visual Basic 6.0 中文版學習範本”。台北市:松崗。 [4]林子傑,李宗寶(2008)。 “利用Multiple Common Vector 及 Dynamic Time Warping 於特定語者中文單音辨識”。碩士論文,國立中興大學應用數學研究所,台中。 [5]張國清,李宗寶(2005)。 “用K-means之動態時間軸校正法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。 [6]楊茗惠,黎自奮,李宗寶(2003)。 “用隱藏式馬可夫方法於頻域特徵之國語數字辨識”。碩士論文,國立中興大學應用數學研究所,台中。 [7]Atal, B.S. and Hanauer, S.L. (1971). “Speech analysis and synthesis by linear prediction of the speech wave”, J. Acoust. Soc. Am. vol. 50, pp. 637-655. [8]Bing, X. and Yihe, S. (1996). “Research on ASIC for multi-speaker isolated word recognition”, ASIC, 2nd International Conference, pp. 135-137. [9]Cover, T.M. and Hart, P.E. (1967). “Nearest Neighbor Pattern Classification”, IEEE Trans. On Information Theory, vol. IT-13, No. 1, pp. 21-27. [10]Cevikalp, H. and Neamtu, M. (2005). “Discriminative Common Vectors for Face Recognition”, IEEE Trans. On Pattern Analysis and Machine Intelligence, vol. 27, No. 1, pp. 4-13. [11]Chu, M.K. and Sohn, Y.S. (2001). “A User Friendly Interface Operated by the Improved DTW Method”, The 10th IEEE International Conference, vol. 3, pp. 1187-1190. [12]Gulmezoglu, M.B., Dzhafarov, V. and Barkana, A. (1999). “A novel approach to isolated word recognition”, IEEE Trans. On Speech and Audio Processing, vol. 7, No. 6, pp. 620-627. [13]Higgins, A.L., Bahler, L.G.. and Porter, J.E. (1993). “Voice identification using nearest-neighbor distance measure”, ICASSP, vol. 2, pp. 375-378. [14]Tran, T.N., Wehrens, R. and Buydens, L.M.C. (2006). “KNN-kernel density-based clustering for high-dimensional multivariate data”, Computational Statistics and Data Analysis, vol. 51, pp. 513-525.

This paper is mainly to discuss the speech recognition of 200 isolated mandarin words for Speaker-Dependent System and we use the method of weighted multiple K-Nearest Neighbor. We consider the three experimental factors such as “the number of frame”, “the proportion of weight” and “the number of K” for the rate of recognition and try to find out the best rate under which kind of combination. By the experimental result, we recognize 200 isolated mandarin words which attain 90.50% for the best rate of recognition.
其他識別: U0005-2906200918245900
Appears in Collections:應用數學系所

Show full item record
TAIR Related Article

Google ScholarTM


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