Query by Humming
Music retrieval; Time series database querying
With the appearance of large scale audio and video databases in various application areas, novel information retrieval methods adapted to the specific characteristics of these data types are required. A natural way of searching in a musical audio database is by humming the tune of a song as a query, which is so-called “query by humming”. In this entry, state-of-the-art techniques for effective and efficient querying by humming are described.
In 1995, Asif Ghias et al.  proposed the basic architecture for a system supporting query by humming. Three main components are introduced in the system: a pitch-tracking module, a melody database, and a query engine. Queries are hummed into a microphone, digitized, and fed into a pitch-tracking module. Then, a symbol sequence representation upon the relative pitch transitions of the hummed melody is sent to the query engine, which produces a ranked list of...
- 1.Fu AW-C, Keogh EJ, Lau LYH, Ratanamahatana CA. Scaling and time warping in time series querying. In: Proceedings of the 31st International Conference on Very Large Data Bases; 2005. p. 649–60.Google Scholar
- 2.Ghias A, Logan J, Chamberlin D, Smith BC. Query by humming: musical information retrieval in an audio database. In: Proceedings of the 3rd ACM International Conference on Multimedia; 1995. p. 231–6.Google Scholar
- 4.Zhou M, Wong MH. Boundary-based lower-bound functions for dynamic time warping and their indexing. In: Proceedings of the 23rd International Conference on Data Engineering; 2007. p. 1307–11.Google Scholar
- 5.Zhu Y, Shasha D. Warping indexes with envelope transforms for query by humming. In: Proceedings of the ACM SIGMOD International Conference on Management of Data; 2003. p. 181–92.Google Scholar