https://github.com/belhoussine/boid

A Neuroevolution simulation that combines both Neural Networks and Genetic Algorithms to play the game Flappy Bird.
https://github.com/belhoussine/boid

boid flappy-bird game genetic-algorithms machine-learning neural-network simulation

Last synced: 6 days ago
JSON representation

A Neuroevolution simulation that combines both Neural Networks and Genetic Algorithms to play the game Flappy Bird.

• Host: GitHub
• URL: https://github.com/belhoussine/boid
• Owner: Belhoussine
• Created: 2020-08-25T22:29:44.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2023-03-01T22:52:35.000Z (over 3 years ago)
• Last Synced: 2025-03-05T12:48:26.725Z (over 1 year ago)
• Topics: boid, flappy-bird, game, genetic-algorithms, machine-learning, neural-network, simulation
• Language: JavaScript
• Homepage: https://boid.belhoussine.com/
• Size: 17.2 MB
• Stars: 1
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Boid

> **- This is a Neuroevolution [simulation](https://boid.belhoussine.com) that combines both Neural Networks and Genetic Algorithms to play the game *Flappy Bird***.  

**- It takes about 20 generations for the boid to become good at the game, and about 50 to master it.**  

**- The simulation speed and birds per generation can be adjusted using the sliders at the bottom left of the page.**  

**- Well performing boids can be saved and loaded into the game.**

> Link: [Boid](https://boid.belhoussine.com)

---------------

![Boid GIF](images/demo.gif)

## Implement a *Flappy Bird* clone

- [x] **Boid class**:

    - [x] Gravity

    - [x] Y-velocity

    - [x] Jump on Key Pressed

    - [x] Die on collision

    - [x] Boid image

    - [ ] Change head pointing position while flying

- [x] **Pipe class**:

    - [x] Two Rectangles

    - [x] X-velocity

    - [x] Spacing 

    - [x] Pipe Image

- [x] **Create Environment**:

    - [x] Create World

    - [x] Create Boid

    - [x] Create Pipes

    - [x] Handle collisions

    - [x] Background Image

    - [x] Make a Scoreboard 

    - [x] Track generations

## Implement a *Neural Network* 

- [x] Design NN Architecture (10/10/2 vs 6/10/2)

- [x] Create Layers

- [x] Randomize weights

- [x] Add activation functions

- [x] Copy Neural Network function

- [x] Predict / Think function

- [x] Save NN of boid with Best Fitness in JSON format

- [x] Load NN from JSON and assign it to a boid

## Implement a *Genetic Algorithm*

- [x] Create a Generation of Boids

- [x] Implement Mutate function

- [ ] Implement a Crossover function

- [x] Copy NN with best fitness

- [x] Create a generation of mutated boids from best NN

- [x] Create a slider for learning speed

- [x] Store boids with best Fitness from every generation

**NOTE**: It is recommended to have at least 100 boids per generation for a faster training.