Data-driven music understanding

"Data-driven music understanding" is an NSF-funded project at LabROSA concerned with extracting information from music audio and discovering deeper patterns and structure within it. This web page is a central hub for materials resulting from this project.

Project reports

• Project Summary and Description from the original NSF proposal
• 2007-8 annual report

Theses

• G. E. Poliner (2008)
  Classification-Based Music Transcription
  Ph.D. Thesis, Columbia University Dept. of Electrical Engineering.
  

Publications

2009

2008

• M. Mandel and D. Ellis (2008) Multiple-Instance Learning For Music Information Retrieval Proc. ISMIR 2008, pp. 577-582, Philadelphia, September 2008.
• J. Devaney and D. Ellis (2008) An Empirical Approach to Studying Intonation Tendencies in Polyphonic Vocal Performances J. Interdisc. Music Studies, vol. 2 no. 1-2, Spring/Fall 2008, pp. 141-156. (16pp)
• M. Mandel and D. Ellis (2008) A Web-based Game for Collecting Music Metadata J. New Music Research, vol. 37 no. 2, pp. 151-165, 2008.
• D. Ellis, C. Cotton, and M. Mandel (2008) Cross-Correlation of Beat-Synchronous Representations for Music Similarity Proc. ICASSP-08 Las Vegas, April 2008, pp. 57-60. See also the talk slides.

2007

• C. Smit and D. Ellis (2007) Solo voice detection via optimal cancelation Proc. IEEE Workshop on Apps. of Sig. Proc. to Acous. and Audio WASPAA-07, Mohonk NY, October 2007, pp. 207-210.
• G. Poliner and D. Ellis (2007) Improving generalization for polyphonic piano transcription Proc. IEEE Workshop on Apps. of Sig. Proc. to Acous. and Audio WASPAA-07, Mohonk NY, October 2007, pp. 86-89.
• D. Ellis (2007) Classifying Music Audio with Timbral and Chroma Features Proc. Int. Conf. on Music Info. Retrieval ISMIR-07 Vienna, Austria, pp. 339-340. (See also the poster I presented at ISMIR-07.)
• M. Mandel and D. Ellis (2007) A Web-Based Game for Collecting Music Metadata Proc. Int. Conf. on Music Info. Retrieval ISMIR-07 Vienna, Austria, pp. 365-366. (See also the 6 page tech. report.)
• J. H. Jensen, D. Ellis, M. G. Christensen, S. H. Jensen (2007) Evaluation Distance Measures Between Gaussian Mixture Models of MFCCs Proc. Int. Conf. on Music Info. Retrieval ISMIR-07 Vienna, Austria, pp. 107-108.

Web materials

• MEAPsoft - software for analyzing and reordering music audio
• Speech and Audio Processing and Recognition:: Class materials (slides, assignments, practicals, demonstrations).
• Automatic piano transcription.
• A baseline artist identification system for the artist20 data in Matlab.
• Artist identification using timbral and chroma features (data and code).
• Automatic chord recognition from music audio (data and code).
• Matlab examples of common audio processing algorithms continually updated.

Acknowledgment

This material is based upon work supported by the National Science Foundation under Grant No. IIS-0713334. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).

Last updated: $Date: 2008/11/03 19:00:27 $
Dan Ellis <dpwe@ee.columbia.edu>