论文信息 - MIREX AUDIO CHORD DETECTION

MIREX AUDIO CHORD DETECTION

This extended abstract briefly introduces the submission to the 2008 Music Information Retrieval Evaluation eXchange in the Audio Chord Detection task. The chord vocabulary we are interested in includes only major and minor triads and ‘N’ for non-chord. To represent a chord for a frame, a traditional 12-dimension PCP vector is used and Hidden Markov Models are used as the classifier. The probability of the distribution of the PCP vector is modeled by a 12-dimension multivariate Gaussian. Also, the parameters of the Gaussian and the state transition matrix are learned through supervised learning. Tests were conducted on Beatles music and the result is shown.

M. Omologo | Maksim Khadkevich

[1] Lawrence R. Rabiner,et al. A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.

[2] Masataka Goto,et al. An Audio-based Real-time Beat Tracking System for Music With or Without Drum-sounds , 2001 .

[3] Daniel P. W. Ellis,et al. Chord Recognition and Segmentation Using EM-trained Hidden Markov Models , 2003 .

[4] Masataka Goto,et al. A chorus-section detecting method for musical audio signals , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..

[5] Mark B. Sandler,et al. Symbolic Representation of Musical Chords: A Proposed Syntax for Text Annotations , 2005, ISMIR.

[6] Geoffroy Peeters,et al. Large-Scale Study of Chord Estimation Algorithms Based on Chroma Representation and HMM , 2007, 2007 International Workshop on Content-Based Multimedia Indexing.

[7] Malcolm Slaney,et al. A Unified System for Chord Transcription and Key Extraction Using Hidden Markov Models , 2007, ISMIR.