Towards the digital music library: tune retrieval from acoustic input

  • Authors:

  • Rodger J. McNab;Lloyd A. Smith;Ian H. Witten;Clare L. Henderson;Sally Jo Cunningham

  • Affiliations:

  • Department of Computer Science, University of Waikato, Hamilton, New Zealand;Department of Computer Science, University of Waikato, Hamilton, New Zealand;Department of Computer Science, University of Waikato, Hamilton, New Zealand;School of Education, University of Waikato, Hamilton, New Zealand;Department of Computer Science, University of Waikato, Hamilton, New Zealand

  • Venue:
  • Proceedings of the first ACM international conference on Digital libraries
  • Year:
  • 1996

Quantified Score

Hi-index 0.02

Visualization

Abstract

Music is traditionally retrieved by title, composer or subjectclassification. It is possible, with current technology, toretrieve music from a database on the basis of a few notes sung orhummed into a microphone. This paper describes the implementationof such a system, and discusses several issues pertaining to musicretrieval. We first describe an interface that transcribes acousticinput into standard music notation. We then analyze string matchingrequirements for ranked retrieval of music and present the resultsof an experiment which tests how accurately people sing well knownmelodies. The performance of several string matching criteria areanalyzed using two folk song databases. Finally, we describe aprototype system which has been developed for retrieval of tunesfrom acoustic input.