Inspiration
Thought of music generation at 2 am, thinking it would be cool to try
What it does
Generates music based on input midis, using short phrases within the music that is input
How we built it
Knowledge of what defines a phrase within a piece, intuition about how to define the start and ends of phrases Machine learning for generation of new phrases (still working on integrating with the current idea) Machine learning for sequences of phrases Git
Challenges we ran into
Organization of code Complexity of Machine Learning and deficit of initial knowledge Time Constraints
Accomplishments that we're proud of
The music that was produced
What we learned
Learned about Machine Learning(RBMs and RNN), Tensorflow, and numpy Interacted with music21
What's next for Autoble
Keep on creating music?
Built With
- music21
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.