Dream On, Sing On

Title
Motivation
Features
What is it?
Future Goals
Thank You

Inspiration

We were inspired to work on this project after our friend struggled to find her karaoke song when we went to karaoke yesterday. It can be incredibly frustrating when you're at karaoke, but you can't find the right song or struggle to know the lyrics. We wanted to create a solution that combines the excitement of karaoke with easy access to lyrics and song separation, ensuring a seamless singing experience.

What it does

Dream On, Sing On is a web app that allows users to separate tracks into vocal and instrumental tracks. Additionally, it displays the lyrics of the song, making it perfect for karaoke sessions.

How we built it

We built Dream On, Sing On using the open-source Spleeter (a machine learning model for source separation). The backend processes the uploaded file, splitting it into vocals and instrumental tracks, and then sends the processed files back to the front end. On the front end, we used React with Vite to create an interactive and responsive user interface, which is inspired by a retro '80s nostalgic vibe.

Challenges we ran into

Audio Processing: We spent a lot of time trying to get Spleeter to work due to the setup being quite tricky.
File Uploads: We did not have enough time to implement the file uploading feature to work with the Spleeter system, but we have created some audio demos with our Python scripts.

Accomplishments that we're proud of

Successfully integrated the Spleeter model for accurate vocal and instrumental separation, providing high-quality outputs.
Created a user-friendly interface that allows easy song uploads and displays lyrics synchronized with the instrumental.
Developed a fully functional web app that supports multiple users and provides a smooth karaoke experience.

What we learned

Audio processing is more complex than anticipated. We thought it would be a simple in-and-out kind of setup, but it ended up taking us a collective 10 hours just to get it to work. We gained a deeper understanding and appreciation for the complexities of backend development.

What's next for Dream On, Sing On

Adding more features: We plan to include more advanced features such as user accounts, playlists, and the ability to save favorite songs.
Expanding audio processing: It currently takes a very long time to process songs like 5 minutes, and we want to make it a quick 10 seconds.
File Upload/YouTube API: We want to give users the ability to select songs directly from YouTube and upload their own files.