Inspiration
This project was inspired by Shengming's recent work with sheet music transcribing when he was trying to transcribing one of his music into sheet music, he ran into a lot of trouble setting up for the python environment and running the tool spleeter, an AI-driven audio track separation tool to aid his work.
What it does
This tool uses an AI model to analyze and attempt to separate different musical instrument into its own audio tracks from a single combined audio file.
How we built it
We used python with flask as http server for backend server and audio separation and React.js for front end. The entire tech stack is sitting under a docker container for easy deployment.
Challenges we ran into
Google Cloud App Engine really doesn't like to run with TensrorFlow.
Accomplishments that we're proud of
We figured out how to deploy docker containers!
What we learned
Program containerization, serverless function, and a lot of DevOps tricks on GCP.
What's next for Transcribe Assistant
We hope to add more metrics when the AI analyzes the source audio file.
Built With
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.