Music Structure Analysis and Its Application to Theme Phrase Extraction

Music is an important component of digital libraries. This paper discusses a digital music library from the information retrieval viewpoint and proposes a method for extracting theme phrases. These are then used to present a shorter version of retrieved music to users. The method consists of two steps, phrase extraction and syntactical classication of segmented fragments of melodies. Phrase extraction is carried out based on a few heuristic rules. We conducted an experiment on the accuracy of phrase extraction using 94 Japanese popular songs and obtained 0.766 recall and 0.786 precision. The syntactical classification is based on a probabilistic syntactical pattern analysis combining classification and syntactical analysis. The proposed method uses a decision tree and a finite state automaton and obtained 0.884 accuracy in theme phrase extraction.

[1]  L. Baum,et al.  An inequality and associated maximization technique in statistical estimation of probabilistic functions of a Markov process , 1972 .

[2]  Brian Christopher Smith,et al.  Query by humming: musical information retrieval in an audio database , 1995, MULTIMEDIA '95.

[3]  David J. Spiegelhalter,et al.  Machine Learning, Neural and Statistical Classification , 2009 .

[4]  Robert Rowe,et al.  Interactive Music Systems: Machine Listening and Composing , 1992 .

[5]  Oliver Günther,et al.  Multidimensional access methods , 1998, CSUR.

[6]  R. Jackendoff,et al.  A Generative Theory of Tonal Music , 1985 .

[7]  Curtis Roads,et al.  The Computer Music Tutorial , 1996 .

[8]  Yoichi Muraoka,et al.  An Audio-based Real-time Beat Tracking System and Its Applications , 1998, ICMC.

[9]  King-Sun Fu,et al.  Stochastic Error-Correcting Syntax Analysis for Recognition of Noisy Patterns , 1977, IEEE Trans. Computers.

[10]  Takeo Kanade,et al.  Video skimming and characterization through the combination of image and language understanding , 1998, Proceedings 1998 IEEE International Workshop on Content-Based Access of Image and Video Database.

[11]  Atsuhiro Takasu,et al.  Phrase based feature extraction for musical information retrieval , 1999, 1999 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing (PACRIM 1999). Conference Proceedings (Cat. No.99CH36368).

[12]  Vladimir D. Gusev,et al.  The Search for Adaptations in Song Melodies , 1997 .

[13]  Alberto Maria Segre,et al.  Programs for Machine Learning , 1994 .

[14]  Thomas G. Szymanski,et al.  A fast algorithm for computing longest common subsequences , 1977, CACM.

[15]  Robert Rowe,et al.  Interactive Music Systems: Machine Listening and Composing , 1992 .

[16]  Robert Rowe,et al.  Machine Listening and Composing with Cypher , 1992 .

[17]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[18]  Yasushi Kiyoki,et al.  An Automatic Metadata Creation Method for Music Data and its Application to Semantic Associative Search , 1998 .