Using a genetic algorithm as a gameplay mechanic to play Chrome Dino game🦖

In the game, we implemented our own genetic algorithm which mimics the process of natural evolution that is selection, mutation and crossover. In the framework, the population of Dino evolves iteratively in each generation to increase the fitness of Dino to jump over all the cacti.

Information

Technical documentation of this project is available at url: https://github.com/prakharsdev/ChromeDino/blob/master/UsingGeneticAlgorithmAsAGamePlayMechanicToPayChromeDinoGame.pdf

Environment

python=3.8 is used

Requirements

pip install -r requirements.txt

Execution (approach 1)

Once you clone the repo, you can execute ChromeDino using below command

python GeneticChromeDino.py

Execution logs (approach 1)

Can be found in file executionlog.txt

Execution (approach 2: NEAT Algorithm)

To bring some intelligence in the game, NEAT Algroithm is used. For the algorithm implementation, NEAT-Python library is used. To try out A.I. learning to play the game, run a command

python NEATChromeDino.py