|
Automatic Content-based Music Recommendation
The system emerged from the technology AudioID which is capable of identifying music titles on the basis of distinct characteristics of the audio signal. SoundsLike relies on the extraction of significant musical features from music files. The extracted features of an unknown piece of music are compared to the characteristics of registered audio files in a database. As a result, the user receives a sorted list of results containing similar music – comparable with Internet search engines.
The special thing about SoundsLike is that it works in an entirely automated manner. In contrast to other recommendation engines, the system needs no manually generated database. Moreover, the user may weight differently the used features, for instance, in order to exclusively search for files with a similar rhythm.
SoundsLike allows the automatic generation of playlists depending on a given reference piece – for example in discotheques, restaurants or suitable for certain occasions. In addition, this method is an innovative way to discover unknown and new music. The technology therefore not only offers Internet music sellers new ways of music recommendation and detection. Furthermore, the SoundsLike technology extends the functionalities of professional asset and content management systems in radio stations and agencies.
The AudioID technology has been developed at the Fraunhofer Institute for Integrated Circuits IIS and the Fraunhofer Institute for Digital Media Technology IDMT. The technology is currently being enhanced under the product name “Mufin MusicFinder”. If you want to learn more about licensing conditions and prices please visit www.mufin.com or contact:
mufin GmbH
Friedrichstraße 200
10117 Berlin
Germany
Phone: +49 (0) 30 / 29392 – 454
Fax: +49 (0)30 / 29392 – 400
info@mufin.com
|
