[ DEPRECATED ] This project uses an old version of ml5.js.
Machine Learning Pong it's an experimental version of the famous computer game Pong that aims to educate users on how machine learning tasks are trained through a fun gaming experience. The user is able to teach, in realtime, a neural network and then play the game in the browser using a set of assigned emojis as reference.
How it works
ML Pong uses the P5 DeepLearn library https://github.com/cvalenzuela/p5-deeplearn-js to train the controllers and predict the output.
Video Demo version runs in 2 different windows browsers and is sending the data through web sockets. (The repo for this project combines the game and the controller in one single page)
Currently version of the game runs the game and the controller at the same page.*
- Clone Repository.
privatedirectory inside of
/serverand add your
- Replace your credentials in the
$ npm install
$ npm start
Project will run at port 3000.