Query by humming

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Query by humming (QbH) is a music retrieval system that branches off the original classification systems of title, artist, composer, and genre. It normally applies to songs or other music with a distinct single theme or melody. The system involves taking a user-hummed melody (input query) and comparing it to an existing database. The system then returns a ranked list of music closest to the input query.

One example of this would be a system involving a portable media player with a built-in microphone that allows for faster searching through media files.

The MPEG-7 standard includes provisions for QbH music searches.

Examples of QbH systems[edit]

SoundHound and Midomi are the only commercially available query by humming services available online at Midomi.com or on the mobile app called SoundHound. Both are powered by the same backend and are capable of recognizing humming and singing as well as recorded tracks. For the singing and humming search, the searchable database is based on Midomi.com's user contributions. Midomi has collected about one million tracks based on user contributions in multiple languages, making it the largest database of its kind by a large margin. The top four languages are: English, Japanese, Chinese and Spanish.

"Musipedia" is an example of a QbH system that uses a variety of input methods such as humming, tapping the keyboard, keyboard search (a virtual piano keyboard), draw notes, and a contour search, using Parsons Code to encode the music pieces.

Tunebot is a music search engine that uses queries from humming, lyrics, and melody. People can contribute to the database and expand the variety of searchable songs. Tunebot also serves as the back-end for a game called Karaoke Callout, in which players' performances are compared by the engine with songs in the database.

External links[edit]

Online demos[edit]

General info and articles[edit]