Study on Intelligent Online Piano Teaching System Based on Deep Learning Recurrent Neural Network Model

Abstract

This study has been conducted to solve the problem of repetitive piano lessons and to bring a personalized experience for each piano learner. The application of deep learning (DL) technology for children's piano teaching has a positive effect on their interest in the subject and improves the teaching quality. Music instruments were identified in the system using an instrument recognition model that was developed using deep learning techniques. It was also utilized to help children learn to play the piano by giving them direction and boosting their excitement for it. The proposed model's ability to recognize and acquire features has been improved. The recurrent neural network (RNN) demonstrated instrument recognition accuracy of 96.4%, and the model's recognition error rate decreased and stabilized as the number of iterations increased. The proposed RNN for musical instruments recognizes instruments by using DL to accurately identify musical properties.