University of Twente Student Theses

Login

Enhancing music genre classification with neural networks by using extracted musical features

Flederus, D.R. (2020) Enhancing music genre classification with neural networks by using extracted musical features.

[img] PDF
3MB
Abstract:The use of mel-frequency cepstral coeffcients (MFCCs) has proven to be a powerful tool in music and voice recognition, and sound recognition in general. This paper is focused on investigating what data we can use along with MFCCs to increase the accuracy of music genre classification. The results of this process are analyzed to gain insight into the characteristics of different music genres. In this paper, MFCCs are considered for music genre classification using a multilayer perceptron (MLP) neural network. Measuring the effect of augmenting MFCCs with additional audio features at the input of the MLP. Following this there is an analysis of the effects different features, e.g. zero-crossing rate, spectral bandwidth etc., have on the accuracy of classifying genres and what the results show about the similarities and relatedness of music genres. Finally, an analysis of the results of classifying a selection of songs from the metal genre.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:http://purl.utwente.nl/essays/80549
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page