Music information retrieval (MIR) technologies are becoming increasingly more important with the growing need for mining and searching for vast amounts of music archive. In addition, song recognition is the method of recognizing a part of a song from either a digital or an analog audio source. This has contributed to the need for songs to be found within a database consisting of a very large number of entries compared to previous years. This paper present a K-modesclustering framework for audio fingerprint database. In addition, introduced the extraction method of the binary encoded fingerprint which take advantage of the K-modes model. Finally, a discussion on the proposed framework with the support evaluation recognition accuracy.
Audio fingerprint database structure using k-modes clustering
Chaouch C.;Sahbudin M. A. B.;Scarpa M.;Serrano S.
2020-01-01
Abstract
Music information retrieval (MIR) technologies are becoming increasingly more important with the growing need for mining and searching for vast amounts of music archive. In addition, song recognition is the method of recognizing a part of a song from either a digital or an analog audio source. This has contributed to the need for songs to be found within a database consisting of a very large number of entries compared to previous years. This paper present a K-modesclustering framework for audio fingerprint database. In addition, introduced the extraction method of the binary encoded fingerprint which take advantage of the K-modes model. Finally, a discussion on the proposed framework with the support evaluation recognition accuracy.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
