We were inspired by being able to leverage AI to create content based on a user's needs. We wanted to explore how AI could help young and coming rappers with their content creation.
What it does
Based on a user's lyrics, we generate a beat, consisting of hi-hats, snares, kick, a melody and a set of chords. This is then given to the user to allow them to rap over this beat.
How we built it
We built the predictive model and beat generator in Python using TensorFlow's magenta library. We also used Google Cloud to analyze speech to text when a user would record into a microphone. The web app was built with HTML and CSS.
Challenges we ran into
We had a hard time choosing the right instruments and making sure the beat generator was generating a beat in sync with the user's lyrical flow.
Accomplishments that we're proud of
We were able to successfully generate a rap beat with all the main components. We were also able to integrate many different many pieces of technology.
What we learned
We learned the benefits of technology on the music industry and how much more there is to explore in this space.
What's next for Beat It
We hope to refine our algorithm to create a wide variety of beats that cater to other content creators who are trying to get their feet of the ground.