複音音樂的音高分析一直是困難的課題之一，在一個時間點內會有多的音高出現，要正確的分辨出是音高或者是諧波不是一件容易的事。
此系統主要是分析複音音訊的音高(pitch)以及諧波(partial)的資訊，主要的目的是希望能夠透過分析音訊結果與其他樂器模型相結合，以便我們用不同的樂器來做合成。此系統包含了尋找複音音樂的基頻以及諧波，並透過隱藏式馬可夫模型(Hidden Markov Model, HMM)將不同音框(frame)的基頻和諧波串連起來。
首先我們利用音樂特性：諧波頻率理論上為基頻倍數，運用最大公因數投票機制來找出基頻的資訊，再透過基頻資訊找出相對應的整個泛音結構，最後透過HMM找出整首音訊的音高與諧波資訊。
有了這些資訊，我們就可以分析出演奏者的手法，進而與其他樂器合成；這樣的作法，相當於用另一種樂器來演奏原來的樂曲，還能保有原來演奏者的手法。

It is one of the most difficult issues to analyze the fundamental frequency (F0) of polyphonic music signal. There are usually many pitches at the same time, so it is hard to recognize each pitch or the corresponding partials correctly.
The main target of this system is to analyze pitches and partials information of polyphonic music signal. We wish to combine the parameters with other musical instrument models to synthesize. This system contains F0 and partial detection, and by Hidden Markov Model (HMM), we can track pitches and partials cross analyzed frames.
This system comprises several parts. First of all, we detect F0s by WGCDV (weighted greatest common divisor and vote) based on WGCDV. Second, we extract the harmonic structure. Finally, we track the pitches and partials of the whole music signal using HMM.
With the above information, we can analyze the styles of the performers and then we may also synthesize using other model of music instrument.

[1] Y. Siao and A. Su, "A new pitch estimation method for musical recordings of solo bowed-string and wind instruments," in Workshop on Computer Music and Audio Technology, Taipei, Taiwan, R.O.C., 2006.
[2] P. Depalle, G. Garcia, X. Rodet, and P. IRCAM, "Tracking of partials for additive sound synthesis using hidden Markov models," in Proc. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP'93), Minneapolis, Minnesota, USA, 1993, pp. 225 - 228.
[3] M. Lagrange, S. Marchand, M. Raspaud, and J. Rault, "Enhanced partial tracking using linear prediction," in Proc. of the 11th Int. Conference on Digital Audio Effects (DAFx-03), London, UK, 2003, pp. 141–146.
[4] M. Lagrange, S. Marchand, and J. Rault, "Using linear prediction to enhance the tracking of partials," in Proc. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP'04), Montreal, Quebec, Canada, 2004, pp. 241–244.
[5] M. Lagrange, S. Marchand, and J. Rault, "Partial tracking based on future trajectories exploration," in 116th Convention of the Audio Engineering Society, Berlin, Germany, 2004.
[6] W.-C. Chang, A. W. Y. Su, C. Yeh, A. Roebel, and X. Rodet, "Multiple-F0 tracking based on a high-order HMM note model," in Proc. of the 11th Int. Conference on Digital Audio Effects (DAFx-08), Espoo, Finland, 2008, pp. 379-386.
[7] D. K. Mellinger, Event formation and separation in musical sound. Ph.D. thesis, Department of Computer Science, Stanford University, 1991.
[8] K. Martin, "A blackboard system for automatic transcription of simple polyphonic music," Massachusetts Institute of Technology Media Laboratory Perceptual Computing Section Technical Report, 1996.
[9] A. Sterian, Model-based segmentation of time-frequency images for musical transcription. Ph.D. thesis, University of Michigan Ann Arbor, MI, 1999.
[10] M. Lagrange and G. Tzanetakis, "Sound source tracking and formation using normalized cuts," in Proc. Int. Conf. on Acoustics, Speech and Signal Processing (ICASSP'07). vol. 1, Honolulu, Hawaii, USA, 2007, pp. I-61-I-64.
[11] M. Wu, D. Wang, and G. Brown, "A multipitch tracking algorithm for noisy speech," IEEE Transactions on Speech and Audio Processing, vol. 11, pp. 229-241, 2003.
[12] M. Ryynanen and A. Klapuri, "Polyphonic music transcription using note event modeling," in Proc IEEE Workshop on Applications of Signal Processing to Audio and Acoustics(WASPAA'05), Mohonk, NY, USA, 2005, pp. 319-322.
[13] E. Vincent, N. Bertin, and R. Badeau, "Harmonic and inharmonic nonnegative matrix factorization for polyphonic pitch transcription," in Proc. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP'08), Las Vegas, Nevada, USA, 2008, pp. 109-112.
[14] X. Wen and M. Sandler, "A partial searching algorithm and its application for polyphonic music transcription," in Proc. The International Society for Music Information Retrieval(ISMIR'05), London, UK, 2005.
[15] S. Godsill and M. Davy, "Bayesian harmonic models for musical pitch estimation and analysis," in Proc. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP'02), Orlando, Florida, USA, 2002, pp. 1769 - 1772.
[16] L. Rabiner and B. Juang, "An introduction to hidden Markov models," IEEE ASSP Magazine, vol. 3, pp. 4-16, 1986.
[17] L. Rabiner, "A tutorial on hidden Markov models and selected applications in speech recognition," Proceedings of the IEEE, vol. 77, pp. 257-286, 1989.
[18] PianoUPM, "http://pianoupm.euitt.upm.es/database.htm."