Deteksi Ritme pada Musik menggunakan Metode Convolutional Neural Network (CNN)

Abstract

Rhythm detection in music has become quite important today. However, manual rhythm detection in music or songs requires significant time and effort, especially when analyzing large amounts of music data. Therefore, the development of automated methods for rhythm detection in music is becoming increasingly essential. One method that has proven effective in analyzing complex data such as music is Convolutional Neural Network (CNN). With CNN, rhythm detection can be performed more accurately and efficiently, resulting in more reliable models for real-world applications. The testing results show that the system is capable of detecting rhythms from music with an accuracy of 92.29% and is also able to measure the level of plagiarism between songs with consistent performance. Nevertheless, the accuracy of the system can still be further improved for future research by expanding and enriching the dataset used in training the CNN model.