music search

(redirected from Music information retrieval)
Also found in: Acronyms, Wikipedia.

music search

To find information about songs and music. There are many music database and identification services on the Internet that enable users to find missing track titles for their CDs and digital files, as well as identify songs by sound, singing or humming. There are also services that recommend songs to users based on their taste (see music recommendation service). Following are the various ways music databases are searched.

Search by Disc Table of Contents
Although added later (see CD Text), the only descriptive data (meta-data) in the original CD music disc were a table of contents (TOC) containing track numbers and lengths. In order to display artist and song titles on the computer or add them to MP3 files when ripping CDs, several databases were developed that identify a disc by table of contents. See music CD identification.

Search by Digital File Fingerprint
MP3, AAC and other compressed audio files contain fields (tags) for the names of the album, artist, tracks, genre, etc. Tags with missing or incorrect data can be updated with the correct data. An acoustic fingerprint of the file is computed and sent to a fingerprint database for matching. See music file identification and acoustic fingerprint.

Search by Audio Fingerprint
Songs can be identified by a sample of the actual sound being played on the radio. An acoustic fingerprint of the sample is computed and sent to a fingerprint database for matching. A fingerprint algorithm for audio is more sophisticated than one used for file recognition (above). The reason is that the audio sample can be a few seconds from any part of the song, whereas a digital file is entirely available for analysis. See Shazam, midomi, Mobile MusicID, MusicDNS and acoustic fingerprint.

Search by Human Example (Sing, Hum, Tap, etc.)
People can find the name of a song or musical piece by singing or humming the tune or tapping out the melody and rhythm. See midomi and Musipedia.
References in periodicals archive ?
However, there is high interest in establishing a systematic, TREC-like paradigm for the music information retrieval research community [2], so that having a ground truth could be very helpful.
Besides such kinds of usual tasks, there are many other music information retrieval problems including searching compositions by their emotional properties (this task appears in recommendation systems), identifying exact particular performance instead of cover song retrieval, locating a position inside a song (used in score following, for instance), and many others (3).
As any music public services librarian knows, keeping track of information at the expression level is one of the biggest challenges facing music information retrieval.
The 2005 Music Information retrieval Evaluation Exchange (MIREX 2005): Preliminary Overview.
13, 2013 /PRNewswire/ -- NQ Mobile (NYSE: NQ), a leading global provider of mobile Internet services, today announced the release of "Music Radar," a revolutionary content-based music information retrieval (MIR) application from one of its subsidiaries, Yinlong.
Frans Wiering (Department of Information and Computing Sciences, Utrecht University) started in the oposite corner of the field, examining the frontiers between information technology and musicology, opening a view to the future of music information retrieval and music xml coding.
This scheme is certainly a prime candidate for a variety of applications in the developing field of music information retrieval, where simple database queries need simple expression, while the transmission of complex scores demand sophisticated representation.
The upshot: more efficient music information retrieval (MIR) research, easier development of commercial music services, and, ultimately, more powerful, accurate predictions of taste in apps for fans and the music industry.
We also had technical sessions on the latest developments in music information retrieval and facetted browsing.
At the beginning of this new century we find a rapidly evolving domain for research and very pragmatic implementation known as Music Information Retrieval (MIR).
The CAML sessions, expertly organized by program cochairs Helene Boucher (Bibliotheque et Archives nationales du Quebec) and Denise Prince (Conservatoire de musique de Quebec), treasurer Brian McMillan (McGill University), and local arrangements coordinator Christiane Melancon (Universite de Montreal), focused on new audio-visual service needs of library users, the technological revolution and the music business, music information retrieval, and performance libraries.
The One Llama team has collaborated with world-class experts in Information Sciences, Semantic Web and Music Information Retrieval in a relentless quest to hone their technology.

Full browser ?