本文以溫尼濾波的概念來實現單一音源分離兩種混合在一起的樂器旋律。系統分為訓練過程以及分離過程。在訓練過程中，萃取各樂器的特徵存入資料庫，對於特徵萃取的方法提出融合法。分離過程則利用訓練過程萃取的各樂器特徵來分離不同樂器的旋律。在這個階段，依不同的混合訊號輸入計算適合此訊號的特徵權重，以調配出最接近原始單一樂器旋律的訊號。
　　實驗中利用鋼琴、小提琴、長笛以及鼓四種樂器的音訊兩兩互相混合，觀察它們在不同特徵數的分離結果，並歸納各樂器的分離特性。實驗中發現，特徵的相關性及數量會影響分離結果，當特徵數太少且特徵相關性高時，特徵能夠描述的樂器特性有限，造成分離後部份音訊喪失的情況。
　　對於分離品質的評量，除了利用SIR來觀測還加入了Segmental SIR以及Local SIR，如此更能客觀的審視分離結果。

　　In this paper, a Wiener-based method is implemented for the single sensor instrument streams separation system.
　　The streams separation system includes two processes: training process and separation process. In training process, features are extracted for different instruments and a database is constructed. About feature extraction, we propose a new method－Merge. In the separation process, features extracted from training process are applied to separate the mixed instrument sources. During this process, we calculate feature weighing according to different input mixed instrument sources.
　　Experiments were conducted to observe separation characteristic of four instruments: piano, violin, flute, and drums using different number of features. And we find that separation results were influenced by correlation of features and number of features.
　　In addition to SIR, segmental SIR and local SIR were also used to measure the separation performance. By using these three methods, we can determine the separation performance in an objective point of view.

Bach, F. R., and Jordan, M. I. , 2005, Blind One-Microphone Speech Separation: A Spectral Learning Approach, Advances in Neural Information Processing Systems (NIPS), 17, 65-72.

Bertsekas, D. P. , 1999, Nonlinear Programming, second edition, MIT.

Bijaoui, A. , 2002, Wavelets, Gaussian Mixtures and Wiener Filtering, Signal Processing, 82, 709-712.

Bonaroya, L., and Bimbot, F. , 2003, Wiener Based Source Separation with HMM/GMM Using a Single Sensor, 4th International Symposium on Independent Component Analysis and Blind Signal Separation, Nara, Japan, 957-961.

Bonaroya, L., Bimbot, F., and Gribonval, R. , 2006, Audio Source Separation with a Single Sensor, IEEE Transactions on Audio, Speech, and Language Processing, 14(1), 191-199.

Bonaroya, L., Bimbot, F., Gravier, G., and Gribonval, R. , 2006, Experiments in Audio Source Separation with one Sensor for Robust Speech Recognition, Speech Communication, 48, 848-854.

Benaroya, L., Donagh, L. M., Bimbot, F., and Gribonval, R. , 2003, Non Negative Sparse Representation for Wiener Based Source Separation with a Single Sensor, IEEE International Conference on Acoustics, Speech, and Signal Processing, 6, 613-616.

Casey, M. A., and Westner, A. , 2000, Separation of Mixed Audio Source by Independent Subspace Analysis, Proceedings of the International Computer Music Conference, 154-161.

Edwin, K. P. Chong, and Stanislaw H. Żak , 2001, An Introduction to Optimization, John Wiley and Sons, Inc.

Essid, S., Richard, G., and David, B., 2006, Instrument Recognition in Polyphonic Music Based on Automatic Taxonomies, IEEE Transactions on Audio, Speech, and Language Processing, 14(1), 68-80.

Gribonval, R., Bonaroya, L., Vincent, E., and Févotte, C. , 2003, Proposals for Performance Measurement in Source Separation, 4th International Symposium on Independent Component Analysis and Blind Signal Separation, Nara, Japan, 715-720.

Hillery, A. D., and Chin, R. T., 1991, Iterative Wiener filters for image restoration, IEEE Transactions on Signal Processing, 39(8), 1892-1899.

Hyvärinen, A., and Oja, E. , 2000, Independent Component Analysis: Algorithms and Applications, Neural Network, 13, 411-430.

Lampropoulos, A. S., Lampropoulou, P. S., and Tsihrintzis, G. A., 2005, A Middleware System for Web-based Digital Music Libraries, Proceedings of the 2005 IEEE/WIC/ACM International Conference on Web Intelligence, 136-142.

Lee, T., Lewicki, M., Girolami, M., and Sejnowski, T. , 1999, Blind Source Separation of More Source Than Mixtures Using Overcomplete Representations, IEEE Signal Processing Letters, 6(4), 87-90.

Makino, S., Araki, S., Mukai, R., and Sawada, H. , 2004, Audio Source Separation Based on Independent Component Analysis, Proceedings of International Symposium on Circuits and Systems, 5, 668-670.

Nachtegael, M., Van der Weken, D., Van De Ville, D., Kerre, E., Philips, W., and Lemahieu, I., 2001, An overview of classical and fuzzy-classical filters for noise reduction, 10th IEEE International Conference on Fuzzy Systems, 1, 3-6.

Ott, L., and Longnecker, M., 2001, An Introduction to Statistical Methods and Data Analysis, Duxbury.

Ozerov, A., Philippe, P., Gribonval, R., and Bimbot, F. , 2005, One Microphone Singing Voice Separation Using Source-Adapted Models, IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, 90-93.

Pardo, B. , 2006, Finding Structure in Audio for Music Information Retrieval, IEEE Signal Processing Magazine, 23(3), 126-132.

Roweis, S. T. , 2000, One Microphone Source Separation, Proceeding of Neural Information Processing Systems, 793-799.

Teddy, S. D., and Lai, E. M.-K., 2004, Model-based Approach to Separating Instrumental Music from Single Track Recordings, 8th International Conference on Control, Automation, Robotics and Vision, 3, 1808-1813.

Tzanetakis, G., and Cook, P., 2002, Musical Genre Classification of Audio Signals, IEEE Transactions on Speech and Audio Processing, 10(5), 293-302.

王小川 , 2005, 語音訊號處理, 全華科技圖書股份有限公司.