2024-03-29T11:15:11Z
https://meral.edu.mm/oai
oai:meral.edu.mm:recid/3346
2021-12-13T00:34:05Z
1582963302567:1597824273898
user-ucsy
Musical Genre Classification using Gaussian Mixture Models
Oo, Su Myat Mon
Aye, Khin San
Digital music is one of the most importantdata types, distributed by the Internet. Automaticmusical genre classification is very useful formusic indexing and retrieval. A method torecognize the genre of music audio is considered.In this paper, the input music is represented withDWT (Discrete Wavelet Transform) coefficientsand classifying the extracted features is performedusing Gaussian Mixture Models (GMM). UsingGMM the optimal class boundaries between fourgroups of genre namely, pop, classic, rock and jazzare obtained. The feature vector from featureextraction step uses wavelet coefficients byhierarchical decomposition as it is easy toimplement as well as it can reduce the computationtime and resources required. Given that GMM is arobust approach that could obtain very goodperformance and a solution based on it ispowerful, the classification is mainly composed ofGMM classifiers. The experimental results indicatethat the proposed approach offer encouragingresults.
2010-12-16
http://hdl.handle.net/20.500.12678/0000003346
https://meral.edu.mm/records/3346