/Chrome-Dinosaur-AI

Teaching AI to play a recreation of the Google Chrome dinosaur game

Primary LanguagePythonMIT LicenseMIT

Chrome Dinosaur AI

Teaching AI to play a recreation of the Google Chrome dinosaur game. The game was made using pygame, and the AI part is done with NEAT-Python. Watch a demo video here

Using the application

  • Clone GitHub repository
  • Download required dependencies: $ pip install -r requirements.txt
  • $ python main.py

How the Neural Network is setup:

The neural network is configured with two inputs, one output, and 0 hidden layers.

  • Inputs: Distance from the dinosaur to the nearest cactus, the dinosaurs y position.
  • Outputs: Jump

I have it setup to start with 15 dinosaurs per generation, however this can be changed by going into the config.txt and changing pop_size. Most of the values in config.txt are default values. You can find a description of all of the settings here

Contribution

Any form of contribution is welcome! Please feel free to fork this project, modify/add to it, and make a pull request!