https://github.com/Gogul09/deep-drum

An interactive drum machine & arpeggio generator created using Google's Magenta JS
https://github.com/Gogul09/deep-drum

deep-learning javascript magenta music rhythm tonejs

Last synced: 8 days ago
JSON representation

An interactive drum machine & arpeggio generator created using Google's Magenta JS

• Host: GitHub
• URL: https://github.com/Gogul09/deep-drum
• Owner: Gogul09
• License: mit
• Created: 2018-09-03T13:30:19.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2019-06-18T08:46:24.000Z (over 5 years ago)
• Last Synced: 2024-08-01T13:33:53.285Z (3 months ago)
• Topics: deep-learning, javascript, magenta, music, rhythm, tonejs
• Language: JavaScript
• Homepage: https://gogul09.github.io/software/deep-drum
• Size: 17.6 MB
• Stars: 120
• Watchers: 9
• Forks: 36
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
![Animation](/assets/deepdrum.gif)

In this interactive demo, I have used Google Magenta's [DrumsRNN](https://github.com/tensorflow/magenta/tree/master/magenta/models/drums_rnn) to generate continuous drum patterns based on your input pattern inside a seed limit and Google Magenta's [ImprovRNN](https://github.com/tensorflow/magenta/tree/master/magenta/models/improv_rnn) to generate arpeggio patterns based on your input notes conditioned on a chord. Based on your seed inputs, the deep neural networks (LSTMs) will generate drum and arpeggio patterns live in your browser!

[Demo](https://gogul09.github.io/software/deep-drum)

[Blog post](https://gogul09.github.io/software/creating-intelligent-music-applications-in-the-browser)

[Google Magenta Demos](https://magenta.tensorflow.org/demos/community/)

[YouTube video instructions](https://www.youtube.com/watch?v=sjo6UlQONLc)