非負矩陣因式分解法的技術可將頻譜分解成一系列的模板向量和能量向量。為了更有效地控制音樂訊號源分離的分解，非負矩陣分解法被延伸加入一些參數化模型和先前的音樂知識。
在分析弓形弦樂器的例子裡，隨著時間變動的參數模型運用和諧音模板取得較優異的結果。因此在我們這項工作上，我們針對小提琴奏鳴曲的情況提出一種綜合框架，它具有樂譜資訊和時間頻譜的參數模型，用此來加以限制非負矩陣分解法。
在本篇論文提出我們所實作的綜合框架在理論以及實驗結果上優於其他同樣是基於非負矩陣分解法的模型。

Techniques based on non-negative matrix factorization (NMF) can decompose a time-frequency magnitude spectrogram into a set of template vectors and activation vectors. To control the decomposition more efficiently on audio source separations, NMF has to be extended using parametric models and prior knowledge.
The time-varying parametric model applied on harmonic templates obtained superior results in the case of analyzing bowed-string instruments. In this work, we exhibit an integrated framework with prior score information and temporal-spectral parametric models to constraint NMF for the cases of violin sonatas.
This thesis presents theoretical and experimental results about the proposed framework that outperform other standard NMF-based models.

[1] D. D. Lee and H. S. Seung, "Learning the parts of objects by non-negative matrix factorization," Nature, vol. 401, pp. 788-791, 1999.
[2] P. Smaragdis and J. C. Brown, "Non-negative matrix factorization for polyphonic music transcription," 2003, pp. 177-180.
[3] P. S. J. Ganseman, and S. Dixon, "Imrpoving PLCA-based score-informed source separation with invertible constant-Q transform," presented at the EUSIPCO 2012, 20th European Signal Processing Conference, Bucharest, Romania, august 27-31, 2012.
[4] P. Smaragdis, B. Raj, and M. Shashanka, "Supervised and semi-supervised separation of sounds from single-channel mixtures," Independent Component Analysis and Signal Separation, pp. 414-421, 2007.
[5] R. Hennequin, R. Badeau, and B. David, "Time-dependent parametric and harmonic templates in non-negative matrix factorization," Proc. of DAFx-10, Graz, Austria, pp. 109-112, 2010.
[6] S. A. Raczyński, N. Ono, and S. Sagayama, "Multipitch analysis with harmonic nonnegative matrix approximation," 2007.
[7] E. Vincent, N. Berlin, and R. Badeau, "Harmonic and inharmonic nonnegative matrix factorization for polyphonic pitch transcription," 2008, pp. 109-112.
[8] W. C. Chang, A. W. Y. Su, C. Yeh, A. Roebel, and X. Rodet, "Multiple-F0 tracking based on a high-order HMM model," 2008.
[9] A. W. Y. Liao, "ANALYSIS AND TRANS-SYNTHESIS OF ACOUSTIC BOWED-STRING INSTRUMENT RECORDINGS–A CASE STUDY USING BACH CELLO SUITES."
[10] S. Ewert and M. Müller, "Using score-informed constraints for NMF-based source separation," 2012.
[11] R. Hennequin, B. David, and R. Badeau, "Score informed audio source separation using a parametric model of non-negative spectrogram," 2011, pp. 45-48.
[12] C. Uhle, C. Dittmar, and T. Sporer, "Extraction of drum tracks from polyphonic music using independent subspace analysis," 2003, pp. 843-847.
[13] T. Joachims, "Making large-Scale SVM Learning Practical.Advances in Kernel Methods - Support Vector Learning, B.Scholkopf and C. Burges and A. Smola(ed.)," MIT-Press, 1999.
[14] J. Woodruff, B. Pardo, and R. Dannenberg, "Remixing stereo music with score-informed source separation," 2006, pp. 314-319.
[15] Piano roll. Available: http://en.wikipedia.org/wiki/Piano_roll
[16] Window function. Available: http://en.wikipedia.org/wiki/Window_function
[17] A. A. Alexandre Galembo, Lola L. Cuddy, and Frank A. Russo, "Perceptual significance of inharmonicity and spectral envelope in the piano bass range," Ryerson University Digital Commons @ Ryerson2004.
[18] E. Vincent, R. Gribonval, and C. Févotte, "Performance measurement in blind audio source separation," Audio, Speech, and Language Processing, IEEE Transactions on, vol. 14, pp. 1462-1469, 2006.