The Application of Convolutional Neural Network Method for Converting Bar Notation Image to MIDI Music Format

Abstract

Musical notation is the science or art of arranging notes or sounds in sequence, combination and temporal relationship to produce a sound composition that has unity and continuity. In Asian countries, especially Indonesia, many musicians have not mastered staff notation. Therefore, it is necessary to design a system that can help translate the image of staff notation into MIDI music format to facilitate learning and reading of staff notation using artificial intelligence. The system that is created can be implemented using the Convolutional Neural Network algorithm to perform image object recognition. Convolutional Neural Network is a deep learning algorithm that is capable of performing independent learning processes for object recognition, object extraction, and classification. Based on the research results, the system built using the Convolutional Neural Network algorithm can read and convert staff notation images with an accuracy of 90%.