Music genres are becoming increasingly diverse, and many people listen to music because it has benefits such as refreshing, motivating or therapeutic. However, with the increasing number of genres, some listeners have a tendency towards the type of genre they like. In Indonesia itself, there are several popular music genres such as pop, folk, rock, indie and dangdut. Classification of music genres is an interesting topic when looking at this behaviour. Several approaches to classify popular music genres include audio and tabular data approaches. In this research, classifying music genres using an image approach by implementing SuperTML to change the form of tabular data into image form, which is then trained using a pre-trained CNN Densenet, succeeded in achieving an accuracy of 67%.
Copyrights © 2023