We love games! We love dancing! We figured it would be an awesome idea to find a way to understand how to process video dances so we can make a game version of it from anything!

What it does

Early progress

When the app is provided with a dance ideo, it will scan that video for a human (preferably one dancing). Then the user will be in front of their webcam, then also scanned. They then will dance, getting a higher score the more accurate your dance is compared to the dance video!

How we built it

Pose Net in Tensorflow did a lot of the heavy lifting, it was created by Google to determine the location of a person's body, their limbs, torso, eyes, nose, ears etc.. We made a web application that utilized this, being able to take any video or gif from any source and create a dancing figure you can compete against. The backend is NodeJS and the front end is ReactJS.

How to run

For package requirements you must run npm install in the Wiggl root dir and the client dir.

Then run in the root dir run node on the server.js file:

node server.js

Then in another terminal, enter the client dir and run via npm

cd client
npm run start

The app will automatically open locally in your web browser. It is located at localhost:3000.

Challenges we ran into

This was a new package for us, and web applications with javascript was completely new to most of us. The project resources were incredibly complicated as well, so it wasn't easy to reverse engineer.

Accomplishments that we're proud of

We managed to make it work!!!!!! We got a dancing figure from a video and learned web application Javascript.

What we learned

Building web applications is insanely complicated, I now understand why the workforce asks for so many constantly.

What's next for Wiggl

We're going to optimze its performance and make the app fully compatible on all devices, and add large player lobbies for dancing! (Ex. 8 or 10 players).

Built With

+ 1 more
Share this project: