Abstract
This research aims to develop a gamelan music genre classifier based on the musical mode system determined based on the dominant notes in a certain order. Only experts can discriminate the musical mode system of compositions. The Feed Forward Neural Networks method was used to classify gamelan compositions into three musical mode systems. The challenge is to recognize the musical mode system of compositions between the initial melody without having to analyze the entire melody using a small amount of data for the dataset. Instead of conducting a melodic extraction from audio signal data, the text-based skeletal melody data, which is a form of extracted melodic features, are used for the dataset. Unique corpuses are controlled based on the cardinality of the one-to-many relationship, and a data mapping technique based on the bars is used to increase the number of corpuses. The results show that the proposed method is suitable to solve the specified problems, where the accuracy in recognizing the class of unseen compositions between the initial melody achieves at 86.7%.
Highlights
MUSIC genre classification is a part of music information retrieval research (MIR) which retrieves information of music to measure musical similarity [1]
This research aims to develop a music genre classification system for Javanese traditional music known as gamelan based on the musical mode system determined based on the dominant notes in a certain order
Notes arrangement based on gamelan rules is melodic patterns used by experts to discriminate the musical mode system of a composition
Summary
MUSIC genre classification is a part of music information retrieval research (MIR) which retrieves information of music to measure musical similarity [1]. Notes arrangement based on gamelan rules is melodic patterns used by experts to discriminate the musical mode system of a composition This became a motivation to develop a gamelan music genre classification system. Khafiizh Hastuti et al / International Journal of Computing, 20(3) 2021, 374-383 segmentation within a short duration as works by [3, 5], the number of corpuses will remain as small as the number of compositions In this condition, the networks will be difficult to learn the patterns in a small dataset. A bar-based data mapping to keep the entire melody being used to train the networks was proposed in order to increase the number of corpuses to solve the small number of corpus problems
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
