本論文提出了一種改進的哼唱檢索系統，能同時使用旋律和歌詞的資訊，以實現更好的性能。首先會進行哼唱分辨，以將「唱」和「哼」分離開來。對於「哼」的查詢，我們套用了只含音高的旋律辨識方法，其在參加MIREX的哼唱檢索項目時被使用，並在該比賽中排名第一。對於「唱」的查詢，我們將旋律辨識和歌詞辨識的分數結合，以利用額外的歌詞資訊。歌詞辨識是基於一個改良過的樹狀網路，此技術常用在語音辨識上。系統整體效能，以錯誤減少率來看，在兩個不同的實驗參數下的前20名之結果中，分別達到 39.01％和23.53％，說明了此系統的可行性。

This paper proposes an improved query by singing/humming (QBSH) system using both melody and lyrics information for achieving better performance. Singing/humming discrimination (SHD) is first performed to distinguish singing from humming queries. For a humming query, we apply a pitch-only melody recognition method that has been used for QBSH task at MIREX with rank-1 per-formance. For a singing query, we combine the scores from melody recognition and lyrics recognition to take advantage of the extra lyrics information. Lyrics recognition is based on a modified tree lexicon that is commonly used in speech recognition. The performance of the overall QBSH system achieves 39.01% and 23.53% error reduction rates, respectively, for top-20 recognition under two experimental settings, indicating the feasibility of the proposed method.

[1] A. J. Ghias, D. C. Logan, and B. C. Smith, “Query by humming-musical information retrieval in an audio database,” in Proc. ACM Multimedia’95, San Francisco, 1995, pp. 216–221.
[2] R. J. McNab, L. A. Smith, I. H. Witten, C. L. Henderson, and S. J. Cunningham, “Toward the digital music library: Tune retrieval from acoustic input,” in Proc. ACM Digital Libraries, 1996, pp. 11–18.
[3] J.-S. R. Jang and M.-Y. Gao, “A query-by-Singing system based on dynamic programming,” in Proc. Int. Workshop Intell. Syst. Resolutions (8th Bellman Continuum), Hsinchu, Taiwan, R.O.C., Dec. 2000, pp. 85–89.
[4] Y.-S. Wu, W.-r. Chu, C.-Y. Chi, D. C. Wu, R. T.-H. Tsai, and J Y.-j Hsu, “The Power of Words: Enhancing Music Mood Estimation with Textual Input of Lyrics,” International Conference on Affective Computing & Intelligent Interaction, pp. 1–6, 2009.
[5] T. Wang, D.-J. Kim, K.-S. Hong, and J.-S. Youn, “Music Information Retrieval System using Lyrics and Melody Information,” Asia-Pacific Conference on Information Processing, pp. 601–604, 2009.
[6] X. Xu, M. Naito, T. Kato, and H. Kawai, “Robust and Fast Lyric Search Based on Phonetic Confusion Matrix,” Proceedings of the International Symposium on Music Information Retrieval, pp. 417–422, 2009.
[7] J.-S. R. Jang, H.-R. Lee, M.-Y. Kao, “Content-based Music Retrieval Using Linear Scaling and Branch-and-Bound Tree search,” in Proc. of IEEE International Conference on Multimedia and Expo, August 2001.
[8] Cambridge University Engineering Department , HTK Web-Site, http://htk.eng.cam.ac.uk/, 2006
[9] AT&T Labs Research , AT&T Labs Research - FSM Library , http://www2.research.att.com/~fsmtools/fsm/, 2008
[10] J.-S. R. Jang, "MIR-QBSH Corpus", MIR Lab, CS Dept, Tsing Hua Univ, Taiwan. Available at the "MIR-QBSH Corpus" link at http://www.cs.nthu.edu.tw/~jang.
[11] J.-C. Chen, J.-S. R. Jang, "TRUES: Tone Recognition Using Extended Segments", ACM Transactions on Asian Language Information Processing, No. 10, Vol. 7, Aug 2008.
[12] MIREX 2009, http://www.music-ir.org/mirexwiki/index.php/Main_Page, 2009
[13] M. Suzuki, T. Hosoya, A. Ito, and S. Makino, “Music Information Retrieval from a Singing Voice Using Lyrics and Melody Information,” EURASIP Journal on Advances in Signal Processing, vol. 2007, Article ID 38727, 8 pages, 2007. doi:10.1155/2007/38727
[14] J.-H Chen, “Content-based Music Emotion Analysis and Recognition”, Master Thesis, CS Dept., National Tsing Hua University, June 2006