跳到主要內容

臺灣博碩士論文加值系統

(35.175.191.36) 您好!臺灣時間:2021/07/30 19:06
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

: 
twitterline
研究生:林明達
研究生(外文):Ta-Ming Lin
論文名稱:基於支援向量機之臺灣蛙類混聲識別系統
論文名稱(外文):Mixed-Class Identification for Taiwan Frog Vocalizations Using Support Vector Machines
指導教授:王駿發雷曉方
指導教授(外文):Jhing-Fa WangSheau-Fang Lei
學位類別:碩士
校院名稱:國立成功大學
系所名稱:電機工程學系碩博士班
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:2009
畢業學年度:97
語文別:英文
論文頁數:50
中文關鍵詞:蛙聲支援向量機動態時軸校正
外文關鍵詞:frog vocalizationDTWSVMs
相關次數:
  • 被引用被引用:0
  • 點閱點閱:167
  • 評分評分:
  • 下載下載:23
  • 收藏至我的研究室書目清單書目收藏:1
近年來,蛙聲辨識在生物學及生態環境監控上被視為一項重要的工具,然而在真實環境中,我們可能聽到不同種類的青蛙於同一時間鳴叫,所以我們需要一個可以辨識不同種類蛙聲混合的系統,本論文提出了一個利用支援向量機(Support vector machines)作為辨識核心的臺灣蛙類混聲辨識系統,我們首先將蛙聲切割成一連串的音節(syllable),每一個音節視為訓練及辨識的基本單位,並利用梅爾倒頻譜係數(Mel-frequency cepstral coefficients)作為音節的特徵參數,再以支援向量機進行訓練及分類,以及投票演算法決定鑑別結果,最後我們將辨識結果與另一個著名的分類器: 動態時軸校正(Dynamic time warping)作為比較。
在我們的實驗中,訓練音節包含了臺灣32種蛙類,測試樣本則由其中的12類於時域上做成2類及3類的蛙聲混合,我們加入了投票機制由已辨識出的音節來決定測試樣本中的青蛙種類,實驗結果證實,我們提出的方法在二類混聲及三類混聲的測試樣本中辨識率分別高達94.70%及88.94%。
Automatic recognition of frog vocalization is considered a valuable tool for biological research and environmental monitor in recent years. In real case, we may hear frog sound with many different types at the same time. So we need an automatic system which can efficiently classify frog spices with mixed different types. This thesis presents a mixed-class identification system for Taiwan frog vocalizations using support vector machines (SVMs). Each frog sound sample is firstly segmented into a succession of syllables which are the basic unit for training and classification. The Mel-frequency cepstral coefficients (MFCCs) are extracted as the syllable features. Then these feature vectors are sent to SVMs classifier for training and classification, and to decide the identification result with voting method. The other well-know classifier, Dynamic time warping (DTW) is also used to compare with SVMs.
In our experiments, training syllables consists of 32 Taiwan frog types and we combine two or three samples within 12 different frog types in time sequence domain to create a mixed-class testing samples. Last, we identify the frog types using voting method by recognized syllables of testing samples. The experimental results show that the proposed method can efficiently promote the average recognition accuracy to 94.70% and 88.94% for two and three types mixed within 12 frog species, respectively.
摘要 I
Abstract II
致謝 IV
Contents V
Index of Figures VII
Index of Tables VIII
Chapter 1 Introduction 1
1.1 Background and Motivation 1
1.2 Related Work 2
1.3 Thesis Organization 2
Chapter 2 The Proposed System 5
2.1 System Overview 5
2.2 Frog Vocalization Preprocessing 6
2.2.1 Syllable Segmentation 6
2.2.2 Feature Extraction 8
Chapter 3 Classification Algorithm 14
3.1 Dynamic Time Warping (DTW) 14
3.1.1 DTW Introduction 14
3.1.2 Dynamic Programming (DP) 14
3.1.3 Restrictions on Warping Function 16
3.1.4 Practical DTW Algorithm 18
3.2 Support Vector Machine (SVM) 19
3.2.1 SVM Introduction 19
3.2.2 Frame Based Classification 27
3.2.3 Multi-Class SVM 28
3.3 Voting Method 29
Chapter 4 Experimental Results 30
4.1 Experiment Setting 30
4.2 Recognition Results for Single-Class Frog Vocalizations 30
4.3 Recognition Results for Multi-Class Frog Vocalizations 34
4.4 The GUI for Frog Vocalizations Identification System 45
Chapter 5 Conclusions and Future Works 47
References 48
[1] 楊懿如, 賞蛙圖鑑: 台灣蛙類野外觀察指南: 中華民國自然與生態攝影學會, 民國87年.
[2] S. Fagerlund, "Bird species recognition using support vector machines," Eurasip Journal on Advances in Signal Processing, 2007.
[3] Hemant Tyagi, R. M. Hegde, H. A. Murthy, and A. Prabhakar, "Automatic identification of bird calls using spectral ensemble average voiceprints," European Signal Processing Conference (EUSIPCO), Sep 2006.
[4] C. Kwan, K. C. Ho, G. Mei, Y. Li, Z. Ren, R. Xu, Y. Zhang, D. Lao, M. Stevenson, V. Stanford, and C. Rochet, "An automated acoustic system to monitor and classify birds," Eurasip Journal on Applied Signal Processing, 2006.
[5] C. J. Huang, Y. J. Yang, D. X. Yang, and Y. J. Chen, "Frog classification using machine learning techniques," Expert Systems with Applications, vol. 36, pp. 3737-3743, Mar 2009.
[6] C. H. Lee, C. H. Chou, C. C. Han, and R. Z. Huang, "Automatic recognition of animal vocalizations using averaged MFCC and linear discriminant analysis," Pattern Recognition Letters, vol. 27, pp. 93-101, Jan 2006.
[7] R. Vergin, D. O'Shaughnessy, and A. Farhat, "Generalized mel frequency cepstral coefficients for large-vocabulary speaker-independent continuous-speech recognition," IEEE Transactions on Speech and Audio Processing, vol. 7, pp. 525-532, Sep 1999.
[8] L. Rabiner and B.-H. Juang, Fundamentals of speech recognition: Prentice Hall PTR, 1993.
[9] H. Sakoe and S. Chiba, "Dynamic programming algorithm optimization for word recognition," IEEE Transactions on Acoustics Speech and Signal Processing, vol. 26, pp. 43-49, 1978.
[10] C. Myers, L. R. Rabiner, and A. E. Rosenberg, "Performance tradeoffs in dynamic
time warping algorithms for isolated word recognition," IEEE Transactions on Acoustics Speech and Signal Processing, vol. 28, pp. 623-635, 1980.
[11] C. J. C. Burges, "A tutorial on support vector machines for pattern recognition," Data Mining and Knowledge Discovery, vol. 2, pp. 121-167, Jun 1998.
[12] C. C. Lin, S. H. Chen, T. K. Truong, and Y. Chang, "Audio classification and categorization based on wavelets and support vector machine," IEEE Transactions on Speech and Audio Processing, vol. 13, pp. 644-651, Sep 2005.
[13] E. Osuna, R. Freund, and F. Girosit, "Training support vector machines: an application to face detection," in Computer Vision and Pattern Recognition, 1997. Proceedings., 1997 IEEE Computer Society Conference on, 1997, pp. 130-136.
[14] A. Ganapathiraju, J. E. Hamaker, and J. Picone, "Applications of support vector machines to speech recognition," IEEE Transactions on Signal Processing, vol. 52, pp. 2348-2355, Aug 2004.
[15] C. C. Chang and C. J. Lin, "LIBSVM: A Library for Support Vector Machines," 2001. Available at http://www.csie.ntu.edu.tw/~cjlin/libsvm.
[16] J. Friendman, Another approach to polychotomous classification, Technical report , Department of statistics, Stanford University, 1996. Available at http://www.-stat.stanford.edu/reports/friendman/poly.ps.Z.
[17] 楊懿如,楊懿如的青蛙學堂, Available at http://www.froghome.idv.tw/index.htm.
[18] 台灣大學動物學系空間生態研究室, 台灣野生動物多媒體資料庫:兩棲類篇, Available at http://learning.froghome.org/D/index.html.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top