automatic music annotation

Since QT/search is limited to text I decided to explore ways to make it fit for searching music as well.

Here's a collection of excerpts on this subject:

http://mill.ucsd.edu/dturnbul/Papers/AutomaticMusicAnnotation2.ppt Automatic Music Annotation
A number of research systems have been developed that automatically annotate music. Automatic annotation uses low-level audio content to describe high-level musical concepts. The four works have been chosen based on their contribution to musical feature design. Although the features have been designed for classification by genre, they are useful for other annotation tasks (emotion, instrumentation, rhythmic structure)…Pachet and Cazaly (2003) review six genre classification systems that have been developed before 2002.

http://www.cs.ucsd.edu/~dturnbul/Papers/annotateMusic.pdf
In the last ten years, computer-based systems have been developed to automatically classify music according to a high-level musical concept such as genre or instrumentation. These automatic music annotation systems are useful for the storage and retrieval of music from a large database of musical content.
In general, a system begins by extracting features for each song. The labels and features for a set of labeled songs are used by a supervised learning algorithm to produce a classifier. This classifier can then be used to provide labels for unlabeled songs. In this paper, we examine commercial and academic approaches to musical annotation involving genre, instrumentation, rhythmic style, and emotion. We also describe various musical feature extraction techniques that have been developed for musical genre classification systems. Lastly, we suggest the use of latent variable models as an alternative to the supervised learning approach for music annotation.
When a listener scans the radio in search of an agreeable station, a decision about whether to settle or to keep searching is made based on a small clip of audio. Immediately the listener can determine if the radio station is broadcasting human speech, music, or silence. In the case of music, he or she can usually understand some notion of genre, instrumentation, and tempo. In some cases, the listener can identify the artist and title of the song if it is similar (or identical) to songs previously heard. The question is whether automatic methods can also be employed to deduce high-level information from audio content.

http://en.wikipedia.org/wiki/Query_by_humming Query by humming
Query by humming (QBH) is a music retrieval system that branches off the original classification systems of title, artist, composer, and genre.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.

http://querybyhum.cs.nyu.edu NYU Query by Humming
A Query by Humming system allows the user to find a song by humming part of the tune.The idea is simple: you hum into the microphone, the computer records the hum and extracts certain features corresponding to the melody and rhythm characteristics, and it then compares the features to the features of the songs in the database.

http://www.wired.com/news/digiwood/0,1412,57317,00.html Wired News: Song Search: A Real Humdinger
It's called "Query by Humming," a type of melody-recognition software program on display at this week's Midem music conference in Cannes that identifies a song by title and composer based on a person humming a few bars into a microphone.

http://querybyhum.cs.nyu.edu/index.php?p=about NYU Query by Humming
A Query by Humming system allows the user to find a song by humming part of the tune.The idea is simple: you hum into the microphone, the computer records the hum and extracts certain features corresponding to the melody and rhythm characteristics, and it then compares the features to the features of the songs in the database.

http://mitpress.mit.edu/catalog/item/default.asp?ttype=6&tid=14057 Computer Music Journal - The MUSART Testbed for Query-By-Humming ...
Online catalogs are already approaching one million songs, so it is important to study new techniques for searching these vast stores of audio. One approach to finding music that has received much attention is Query-by-Humming (QBH).This approach enables users to retrieve songs and information about them by singing, humming, or whistling a melodic fragment. In QBH systems, the query is a digital audio recording of the user, and the ultimate target is a complete digital audio recording.

http://portal.acm.org/citation.cfm?id=354520 A practical query-by-humming system for a large music database
A practical query-by-humming system for a large music database
The most significant improvement our system has over general query-by-humming systems is that all processing of musical information is done based on beats instead of notes. This type of query processing is robust against queries generated from erroneous input. In addition, acoustic information is transcribed and converted into relative intervals and is used for making feature vectors.

http://en.wikipedia.org/wiki/Music_information_retrieval Music information retrieval
Music information retrieval or MIR is the interdisciplinary science of retrieving information from music.

http://www.music-ir.org/research_home.html Music Information Retrieval Research Bibliography homepage
Research and development in Music Information Retrieval (MIR) is such a multi-disciplinary area that we have difficulty keeping track of the work being done in all the disciplines.With this "Music Information Retrieval Annotated Bibliography", we hope to enable the much needed communication across the disciplinary boundaries by providing a) bibliographic control of the MIR literature and b) an important mechanism for members of each discipline to comprehend the contributions of the other disciplines.With the tools provided here, we are able to uniformly represent the important scholarly papers found in the MIR fields. By promoting such communication, MIR will be in a better position to fully realize the benefits that a multi-disciplinary research and development community offers.
http://ismir2001.ismir.net International Symposium on Music Information Retrieval (MUSIC IR 2001)
Interest in music information retrieval (music IR) is exploding.This is not surprising: music IR has the potential for a wide variety of applications in the educational and academic domains as well as for entertainment. Yet, until now, there has been no established forum specifically for people studying music retrieval.