LSTM based approach for Generating Music from MIDI notes

  • Beschi Raja J, Goutham Vignesh S, G.Sandhya, V.Roopa, S.Sampeter

Abstract

The aim of the work is to generate music automatically using Recurrent Neural Network (RNN). A non-expert can also generate a decent quality music using RNN. Everybody likes to listen to music and if there is some way to generate music automatically, particularly decent quality music then it's a big leap in the world of the music industry. Using the existing music data the model was trained. The model has to learn the patterns in music that humans enjoy. Once it learns this, the model will be able to generate music for the user. It will understand the patterns of music to generate new music. The input to the model is a sequence of musical events/notes or MIDI files (Musical Instrumental Digital Interface). This model takes the MIDI files and encode them as inputs for the neural network. In this system, Beethoven’s music style has been selected to train the network to make it able to generate similar music patterns. The output will be the newly generated series of raw music. The proposed system acquired the pattern of musical styles and produce new music.

Published
2020-05-14
How to Cite
Beschi Raja J, Goutham Vignesh S, G.Sandhya, V.Roopa, S.Sampeter. (2020). LSTM based approach for Generating Music from MIDI notes. International Journal of Advanced Science and Technology, 29(9s), 2745 - 2752. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/15441