In this paper, we have explored the use of artificial neural networks (ANNs) for automatically detecting the genre of music. The challenge faced when hand-classifying music is that it is highly dependent on the accuracy and experience of the person classifying it. The main objective of this paper is to build an arrangement which will reduce the burden and increase the accuracy of classifying the genre of music. We use MFCC as feature vectors and use multilayer perceptrons (MLPs) to classify the data into the various genres. We have trained our model on a novel dataset that reflects current trends in music and addresses the problems faced with existing datasets.
展开▼