1 2 1, 3 1 Design and Implementation on a System for Learning Songs by Presenting Musical Structures based on Phrase Similarity Yuma Ito, 1 Yoshinari Takegawa, 2 Tsutomu Terada 1, 3 and Masahiko Tsukamoto 1 Players on musical instruments usually memorize musical scores for concerts and live performances. However, for memorizing songs, it efforts on players so much by playing and listening the song over again. Therefore, the goal of our study is to construct a system for memorizing musical scores based on phrase similarity. The proposed system calculates phrase similarity in the target song, and presents the musical structures and the difference points in similar phrases based on the phrase similarity. The user can understand the musical structure immediately, and memorizing the musical score in the short time because of the reducing duplicated information of the song between the similar phrases. We confirmed that our method had advantages compared with conventional musical scores from the result of evaluation. 1. A B A A A 1 2 3 4 5 1 Graduate School of Engineering, Kobe University 2 Organization of Advanced Science and Technology, Kobe University 3 PRESTO, Japan Science and Technology Agency 1 c 2011 Information Processing Society of Japan
2. 2.1 1) 3) 4),5) ScoreIlluminator 6) 7) 2 2.2 8) ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) 5 (4) (1) (2) (1) (2) (4) (1) (4) (3) (2) 1 9) 1 3. 2 A B 1 2 3.1 Lerdahl GTTM 10) 11) 2 2 c 2011 Information Processing Society of Japan
(a) (b) (a) (b) 4 5 6 DP 2 (a) (b) : 5 2 (a) (b) 3 : 3 3 : 4 DP(Dynamic Programming) 12) N i j d m (i, j) (1) d m (i, j) = w t d t (i, j) + w n d n (i, j) + w p d p (i, j) (1) (i = 1,..., N)(j = 1,..., N) d t (i, j) d n (i, j) d p (i, j) DP w t w n w p DP 2 5 1 1 6 2 5 1 2 3 c 2011 Information Processing Society of Japan
1 1 17 d m (i, j) 7 2 5 tab 3 4 (b) 3 4 (a) 3 4 DP i j d p (i, j) DP d p (i, j) 3.2 8 8(a) 8(b) 2 8(a) 1 8(b) 2 8(b) 8(b) 8(b) tab 7 8(b) 8(b) ( 1 ) 8(b) ( 2 ) 2 (1) (2) ( 3 ) 4 c 2011 Information Processing Society of Japan
(a) (b) 8 ( 4 ) 8(b) ( 5 ) 8(b) (4) 4. 3 XML 3 GTTM GTTM PC Windows 7 Microsoft Visual C# 2008 4 2 2 5 c 2011 Information Processing Society of Japan
1 A B C D 1 11 21 19 16 1.00 2 16 20 20 25 1.03 3 25 19 26 22 1.61 4 25 35 27 40 2.91 4 Crazy Little Thing Called Love Freddie Mercury Headlong Queen Hammer to Fall Brian May I Want It All Queen 4 6 9 A B C D 4 1 a n (n = 1, 2, 3, 4 ID ) a n 2 A B C D p 1 2 3 4 0.71 1.78 0.80 1.74 1.64 1.02 1.62 2.09 0.85 0.86 1.74 0.56 0.55 0.86 0.63 0.57 0.67 1.69 1.43 0.71 0.09 0.07 0.51 0.15 p 0.004 0.0002 0.12 0.02 1.26 0.5 0.22 1.59 0.38 0.04 1.00 0.44 0.50 0.65 0.57 0.01 1 C D C D 2 4 a 1 = 16 19 a2 = 25 27 a4 a n a n 1 A 1 2 x x 16 = 1.00 1.03 x = 15.6 1 11 11 15.6 = 0.71 30 2 2 3 t p 3 p t 1.00 6 c 2011 Information Processing Society of Japan
1.00 2 A 33 p 0.004 A 1 D A A B C B p 4 p 4 2 2 p 2 1 A B 1 A B 3 3 2 5. ( ) (A)(20240009) (B)(21700198) 1) E. Pampalk, S. Dixon, and G. Widmer: Exploring Music Collections by Browsing Different Views, Computer Music Journal, Vol.28, No.2, pp.49 62 (2004). 2) M. Torrens, P. Hertzog, and J. Lluís Arcos: Visualizing and Exploring Personal Music Libraries, Proc. of the 5th International Conference on Music Information Retrieval (ISMIR2004), pp.421 424 (2004). 3) P. Lamere, and D. Eck: Using 3D Visualization to Explore and Discover Music, Proc. of the 8th International Conference on Music Information Retrieval (IS- MIR2007), pp.167 170 (2004). 4) SmartMusic: Vol.44, No.11, pp.2737 2747 (2003). 5) 2007 pp.145 148 (2007). 6) ScoreIlluminator: Vol.50, No.12, pp.2937 2948 (2009). 7) 2007-MUS-69-10 Vol.2007, No.15, pp.41 44 (2007). 8) No.0306 (2003). 9) R. Snyder: Music and Memory, The MIT Press (2001). 7 c 2011 Information Processing Society of Japan
(2003). 10) F. Lerdahl, and R. Jackendolf: A Generative Theory of Tonal Music, The MIT Press, Cambridge (1983). 11),, GTTM,, Vol.48, No.1, pp.284 299 (2008). 12) Vol.27, No.9, pp.483 490 (1971). 8 c 2011 Information Processing Society of Japan