(3.238.235.155) 您好!臺灣時間:2021/05/16 17:41
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:李建德
研究生(外文):Jain-De Lee
論文名稱:稀疏性雙反旋積非負矩陣分解法結合遮罩應用於混合蛙鳴訊號分離之研究
論文名稱(外文):Study on Separating Mixed Frog Sounds Using Sparse Non-negative Matrix Factor 2-D Deconvolution Combined With Mask
指導教授:陳文平陳文平引用關係
指導教授(外文):Wen-Ping Chen
學位類別:碩士
校院名稱:國立高雄應用科技大學
系所名稱:電機工程系
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:101
畢業學年度:100
語文別:中文
論文頁數:90
中文關鍵詞:單通道盲訊號分離非負矩陣分解法遮罩
外文關鍵詞:Single Channel Blind Source SeparationNon-negative Matrix FactorizationMask
相關次數:
  • 被引用被引用:1
  • 點閱點閱:821
  • 評分評分:
  • 下載下載:33
  • 收藏至我的研究室書目清單書目收藏:3
利用聲紋辨識進行野外生物物種調查是近年來生態學家重要的研究之ㄧ。然而野外錄製的聲景資料音檔中,時常會出現異種生物鳴叫的混音聲,進而影響聲紋辨識效果,因此這些混音檔必須進行盲訊號分離,以提高系統的辨識結果。
本論文利用稀疏雙反旋積非負矩陣分解法結合遮罩進行單通道的盲訊號分離,首先以稀疏雙反旋積非負矩陣分解法對混合音檔進行分析,再以分析後的訊號當做遮罩訊號,對遮罩訊號進行修改並且萃取以及修正分離訊號,藉以提升訊號分離後的品質。本研究對八種蛙類進行七種的混音項目,並對七種混音項目進行盲訊號分離以及聲紋辨識。實驗將分離後之訊號,各別切出410個音節,然後進行聲紋辨識實驗,經實驗後發現,稀疏雙反旋積非負矩陣分解法與本論文所提出之方法所分離的訊號之平均辨識率分別為49.43%與79.27%,結果證明本方法能有效提升分離訊號的品質以提升辨識系統的正確率。
The technology which ecologists use voiceprint recognition to investigate the species of wild animals is one of important researches in recent years. However, the recordings made from wilderness are often mixed up with some biophony of other animals and the sounds have influences upon the effects of Voiceprint recognition. Thus, the mixed recordings must be dealt with blind source separation to advance the results of voiceprint recognition.
The thesis integrates sparse non-negative matrix factor 2D deconvolution(SNMF2D) with mask to achieve single channel blind source separation. First step is to analyze the mixed recordings by sparse non-negative matrix factor 2D deconvolution. Then, regard the signals which are analyzed as mask signals, revise the mask signals and extract and improve the separated signals to boost the qualities of the separated signals. The research sets up 7 items of mixed recordings in connection with 8 kinds of frogs and implements them into single channel blind source separation and voiceprint recognition. The experiment splits the separated signals into 410 syllables individually and deals them with voiceprint recognition. The consequence is revealed that the average recognition rate of sparse non-negative matrix factor 2D deconvolution and the method proposed in the thesis are 49.43% and 79.27% respectively. It proved that the method can promote the qualities of separated signals effectively to increase the recognition rate of recognition system.
目錄
中文摘要 ----------------------------------------------------------------------------- i
英文摘要 ----------------------------------------------------------------------------- ii
致謝 ----------------------------------------------------------------------------- iii
目錄 ----------------------------------------------------------------------------- iv
表目錄 ----------------------------------------------------------------------------- vi
圖目錄 ----------------------------------------------------------------------------- vii
一、 緒論------------------------------------------------------------------------------ 1
1.1 研究動機與目標-------------------------------------------------------- 1
1.2 研究貢獻----------------------------------------------------------------- 2
1.3 論文架構----------------------------------------------------------------- 3
二、 相關背景與文獻回顧--------------------------------------------------------- 4
2.1 盲訊號分離-------------------------------------------------------------- 4
2.1.1 雞尾酒會問題-------------------------------------------------- 5
2.2 獨立成分分析法-------------------------------------------------------- 9
2.2.1 基本定理-------------------------------------------------------- 10
2.2.2 置中化與白色化----------------------------------------------- 12
2.2.3 中央極限定理-------------------------------------------------- 17
2.2.4 非高斯量測----------------------------------------------------- 18
2.3 非負矩陣分解法-------------------------------------------------------- 23
2.3.1 基本定理-------------------------------------------------------- 24
2.3.2 目標函數-------------------------------------------------------- 25
2.3.3 更新規則-------------------------------------------------------- 26
2.4 稀疏雙反旋積非負矩陣分解法-------------------------------------- 30
2.4.1 雙反旋積非負矩陣分解法----------------------------------- 30
2.4.2 稀疏編碼與稀疏化-------------------------------------------- 33
2.4.3 稀疏雙反旋積非負矩陣分解法----------------------------- 34
2.5 文獻回顧----------------------------------------------------------------- 39
三、 研究方法------------------------------------------------------------------------ 42
3.1 訊號前處理-------------------------------------------------------------- 43
3.2 分析訊號----------------------------------------------------------------- 46
3.3 遮罩修正----------------------------------------------------------------- 49
3.3.1 遮罩二元化----------------------------------------------------- 49
3.3.2 萃取訊號-------------------------------------------------------- 53
3.3.3 計算混合比例-------------------------------------------------- 54
3.3.4 訊號修正-------------------------------------------------------- 54
3.4 訊號後處理-------------------------------------------------------------- 56
四、 實驗結果------------------------------------------------------------------------ 59
4.1 訊號分離實驗----------------------------------------------------------- 60
4.2 辨識實驗----------------------------------------------------------------- 70
五、 結論與未來展望--------------------------------------------------------------- 72
5.1 結論----------------------------------------------------------------------- 72
5.2 未來展望----------------------------------------------------------------- 73
參考文獻 ----------------------------------------------------------------------------- 74
附錄一 ----------------------------------------------------------------------------- 81
[1]S. Choi, A. Cichocki, H. M. Park and S. Y. Lee,“Blind Source Separation and Independent Component Analysis: A Review”, Neural Information Processing Letters and Reviews, Vol.6, No.1,pp.1-57, Jan. 2005.
[2]余先川,胡丹,盲源分離理論與應用,科學出版社,2011。
[3]J. Bylund, “A humble attempt to solve the cocktail-party problem using blind source separation”Physics Scientific Computing, Vol. 4, No.6, pp.1-16, 2001.
[4]J.-F. Cardoso,“Blind signal separation: statistical principles”, Proceedings of the IEEE, Vol. 86, No. 10, pp. 2009-2025, 1998.
[5]S. Haykin and Z. Chen,“The Cocktail Party Problem”, Neural Computation, Vol. 17, No. 9, pp.1875-1902, Sep. 2005.
[6]A. Hyvärinen and E. Oja,“Independent Component Analysis: Algorithm and Application”, Neural Networks, Vol. 13, No.4-5, pp. 411-430, 2000.
[7]A. Hyvärinen,“Fast and Robust Fixed-Point Algorithms for Independent Component Analysis”, IEEE Transactions on Neural Networks, Vol.10, No.3, pp.626-634, 1999.
[8]張嘉芳,以FastICA 為基礎之時域聲音分離演算法,碩士論文,2003。
[9]陳彥名,以獨立成分分析法萃取背景噪音中語音訊號之研究--於助聽器可能之應用,碩士論文,2004。
[10]J. Herault and C. Jutten,“Space or time adaptive signal processing by neural network models”Proceedings of AIP Neural Networks for Computing, Vol.151, pp.207-211, 1986.
[11]E. Sorouchyari,“Blind separation of sources, part III: Stability analysis”, Signal Processing, Vol. 24, No. 1, pp. 21–29, July 1991.
[12]P. Comon,“Independent Component Analysis, a new concept ”, Signal Processing, Vol. 36, No. 3, pp.287-314, April 1994.
[13]Y. Deville, “Analysis of the convergence properties of self-normalized source separation neural networks”IEEE Transactions on Signal Processing, Vol. 47 , No. 5, pp. 1272-1287, May 1999.
[14]R. Linsker,“Self-organization in a perceptual network”, IEEE Computer, Vol. 21, No.3, pp.105-117, 1988.
[15]R. Linsker,“An application of the principle of maximum information preservation to linear systems”, Advances in neural information processing systems , pp.186-194, 1989.
[16]J. Karhunen,“A class of neural networks for independent component analysis” , IEEE Transactions on Neural Networks, Vol. 8 , No. 3, pp. 486-504, May 1997.
[17]E. Oja,“The nonlinear PCA learning rule in independent component analysis”, Neurocomputing, Vol. 17, No. 1, pp. 25-45, Sep. 1997.
[18]H. H. Yang and S.-I. Amari,“Adaptive online learning algorithms for blind separation: maximum entropy and minimum mutual information”, Neural Computation , Vol. 9, No. 7, pp. 1457-1482, Oct. 1997.
[19]A. J. Bell and T. J. Sejnowski,“An information-maximization approach to blind separation and blind deconvolution ”, Neural Computation, Vol. 7, No. 6, pp. 1004-1034, 1995.
[20]T. W. Lee, Independent Component Analysis: Theory and Applications, Kluwer Academic Publishers, Boston, 1998.
[21]T. M. Cover and J. A. Thomas, Elements of Information Theory, Wiley ,1991.
[22]K. Pearson, “On Lines and Planes of Closest Fit to Systems of Points in Space”, Philosophical Magazine, Vol. 2, No.6 , pp. 559-572, 1901.
[23]H. Hotelling,“Analysis of a Complex of Statistical Variables into Principal Components”, Journal of Education Psychology, Vol. 24, No.6 ,pp.417-441, 1933.
[24]P. Paatero and U. Tapper,“Positive matrix factorization: A non-negative factor model with optimal utilization of error estimates of data values”,Environmetrics, Vol.5, No.2, pp. 111-126, June 1994 .
[25]D. D. Lee and H. S. Seung,“Learning the parts of objects by non-negative matrix factorization”, Nature, Vol.401, pp. 788-791, 1999.
[26]D. D. Lee and H. S. Seung,“Algorithms for Non-negative Matrix Factorization”, Advances in Neural Information Processing Systems, Vol.13, pp.556-562, 2001.
[27]M. D. Plumbley, S. A. Abdallah, J. P. Bello, M. E. Davies, G, Monti and M. B. Sandler,“Automatic music transcription and audio source separation”, Cybernetics and Systems, Vol.33, No.6, pp. 603-627, 2002.
[28]P. Smaragdis and J. C. Brown,“Non-Negative Matrix Factorization for Polyphonic Music Transcription”, IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, New Paltz, NY, Oct. 2003.
[29]P. Sajda, S. Du, T. R. Brown, R. Stoyanova, D. C. Shungu, X. Mao and L. C. Parra,“Non-Negative Matrix Factorization for Rapid Recovery of Constituent Spectra in Magnetic Resonance Chemical Shift Imaging of The Brain”,IEEE Transactions on Medical Imaging, Vol.23, pp.1453-1465, Dec. 2004.
[30]D. Guillamet, B. Schiele and J. Vitria“Analyzing Non-Negative Matrix Factorization For Image Classification”, 2002 International Conference on Pattern Recognition, Vol.2, pp.116-119, Aug. 2002.
[31]Z. Junying, L. Wei, Q. Miao and Y. Wang,“Image Fusion based on Non-Negative Matrix Factorization”,2004 International Conference on Image Processing, Vol.2, pp.973-976, Oct. 2004.
[32]J. H. Chou, Optimization Approaches Lecture Notes and Handouts at the Institute of Electrical Engineering ,National Kaohsiung University of Applied Sciences, Taiwan, 2011.
[33]M. N. Schmidt and M. Mørup,“Nonnegative Matrix Factor 2-D Deconvolution for Blind Single Channel Source Separation”, Proceedings of International Conferences Independent Component Analysis and Blind Signal Separation, Vol.3889, Charleston, USA, pp. 700-707, Mar. 2006.
[34]P. O. Hoyer,“Non-negative matrix factorization with sparseness constrains”, Journal of Machine Learning Research, Vol. 5, pp. 1457-1469, 2004.
[35]J. Eggert and E.korner,“Sparse coding and nmf”, Neural Networks, Vol.4, pp.2529-2533, 2004.
[36]M. Morup and M. N. Schmidt,“Sparse non-negative matrix factor 2-D deconvolution”, Technical University of Denmark, Copenhagen, Denmark, 2006.
[37]M. A. Casey and A. Westner,“Separation of Mixed Audio Sources by Independent Subspace Analysis”, Proceedings of the International Computer Music Conference, Berlin, Germany, pp. 154-161, Aug. 2000.
[38]Md. K. I. Molla and K. Hirose,“Single-Mixture Audio Source Separation by Subspace Decomposition of Hilbert Spectrum”, IEEE Transactions on Audio, Speech, and Language Processing, Vol. 15, No. 3, pp. 893-900, Mar. 2007.
[39]L. Bonaroya and F. Bimbot,“Wiener Based Source Separation with HMM/GMM Using a Single SensorHMM/GMM Using a Single Sensor”, International Symposium on Independent Component Analysis and Blind Signal Separation, Nara, Japan, pp.957-961, 2003.
[40]E. M. Grais and H. Erdogan,“Single Channel Speech Music Separation Using Non-Negative Matrix Factorization and Spectral Masks”, IEEE Digital Signal Processing, Sedona, Arizona, Jan. 2011.
[41]P. Smaragdis,“Non-negative Matrix Factor Deconvolution; Extracation of Multiple Sound Sources from Monophonic Inputs”, International Symposium on Independent Component Analysis and Blind Source Separation ,Vol.3195, pp.494-499, 2004.
[42]J. Lin and A. Zhang,“Fault feature separation using wavelet-ICA filter”, NDT & E International, Vol. 38, No. 6, pp. 421-427, 2005.
[43]M. E. Davies and C. J. James,“Source separation using single channel ICA”, Signal Processing, Vol. 87, No. 8, pp. 1819-1832, 2007.
[44]B. Mijović, M. D. Vos, I. Gligorijević, J. Taelman and S. V. Huffel,“Source Separation From Single-Channel Recordings by Combining Empirical-Mode Decomposition and Independent Component Analysis”, IEEE Transactions on Biomedical Engineering, Vol.57, No.9, Sep. 2010.
[45]X. Cheng, N. Li, Y. Cheng and Z. Chen,“A Single-channel Mix Signal Separation Technique”, International Conference on Bioinformatics and Biomedical Engineering, pp. 948 -951, July 2007.
[46]J. F. Cardoso and A. Souloumiac,“Blind beamforming for non-Gaussian Signals”, IEEE Proceedings, Vol. 140, No. 6, pp. 362-370, Dec. 1993.
[47]Z. Y. Liu, K. C. Chiu and L. Xu,“One-Bit-Matching Conjecture for Independent Component Analysis”, Neural Computation, pp. 383-399, 2004.
[48]D. Gabor,“theory of communication”, Journal. Institute of Electrical Engineers, Vol. 93, No.3 ,pp.429-457, 1946.
[49]K. W. Wilson, B. Raj, P. Smaragdis and A. Divakaran,“Speech denoising using nonnegative matrix factorization with priors”, Proceedings of International Conference on Acoustics , Speech, and Signal Processing, pp. 4029 - 4032, 2008.
[50]P. S. Liao, T. S. Chen and P. C. Chung,“A Fast Algorithm for Multilevel Thresholding”, Journal of Information Science and Engineering, Vol.17, pp.713-737, 2001.
[51]National Instruments LabView Webpage [Online], Available: http://www.ni.com/labview/
[52]Jeffrey Travis, Jim Kring, 喬瑞萍 等譯,Labview大學實用教程(第三版),電子工業出版社,2008。
[53]E. Vincent, R. Gribonval and C. Févotte,“Performance measurement blind audio source separation”, IEEE Trans. Speech Audio Process, Vol. 14, No.4, pp. 1462-1469, Jul. 2005.
[54]W. P. Chen, S. S. Chen, C. C. Lin, Y. Z. Chen and W. C. Lin,“Automatic recognition of frog calls using a multi-stage average spectrum”, Computers and Mathematics with Applications, Jan. 2012.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top