Automatic music generation using RNN
V. Bhuvana Kumar, M. Kathiravan · 2023
this paper explains the development of music automatically. Deep learning is employed in order to do this. In the shape of an ABC note sequence, melody is created. Using vast amounts of data for music is presently feasible. mostly for the creation of music utilizing deep learning For modeling, it utilize LSTM or GRU. In terms of music creation, it is comparable to sequence creation. Since LSTM creates sequences most effectively, using it is recommended. Subjective evaluation is carried out by people with diverse backgrounds, including classical music interest, performance, composer, and digital composer. The results reveal that the double stacked layer LSTM model outperforms the composer pattern in music, with a 70% recall score. Furthermore, subjective evaluation indicates that the created music is listenable and intriguing, with a score of 6.85 out of 10 on double LSTM stacked layer.