IO | Devpost

our main page

Inspiration

In our car ride to the hackathon, Vaibhav was attempting to sing songs and we couldn't bear to listen to it. And so in order to lessen society's sufferings we decided to develop an application to help Vaibhav and people like him sound the best they can.

What it does

This application takes an input of a user's singing video, makes the voice clearer and more pleasant using autotune, adds catchy music and beats synced up with the music, making each video unique to the user's voice. Additionally, it provides feedback for the user to get better at singing naturally and offers the option to remove background noise for clarity.

How we built it

We built IO using designing our frontend using NextJS. For our backend, we used a combination of machine learning tools, algorithms, and flask.

Challenges we ran into

It was difficult to sync the beats of the music with the singer's voice, especially with their tonality changed throughout the video.

Accomplishments that we're proud of

We are extremely proud of being able to generate entire musical segments based off a voice recording. We are also proud that each user can make a video unique and tailored to their own voice, rather than being limited by certain options. We are also very proud of our ionic mode which combines multiple instruments together to create our superior beats and soundtracks.

What we learned

We learned that AI can boost creative thinking rather than hinder it through the creation of this project.

What's next for IO

We plan to bring IO to a larger audience by deploying it publicly and refining the features. Additionally, we would like to give our users the option to make music beyond simply "rap" and "chill".