Genetic Flap

Birds learn to beat the game with NN.

Team
Solo
Engine
Unity
Type
Technical study
Status
Prototype

What I built

  • Genetic Algorithm Mechanics: Birds evolve through generations based on their performance, learning to navigate the course more effectively over time.
  • Dynamic 2D Navigation: The game utilizes raycasting and collision detection to simulate realistic bird movements and interactions with obstacles.
  • Fitness-Based Evolution: Birds with better navigation skills are selected for reproduction, enhancing their offspring's chances of survival.
  • Object-Oriented Code Structure: The game is built using clean, object-oriented principles, with distinct classes for birds, populations, and genetic algorithms.
  • Unity Debugging Tools: Real-time visualization of raycasts and collisions to debug and refine the game mechanics.
  • Progressive Difficulty: As the game evolves, new obstacles and challenges can be added to increase difficulty and complexity.

Gameplay

Screenshots

Built with

Engine
Unity
Languages
C#
Tools
Git
Disciplines
Genetic algorithmsMachine learning2D collision detectionRaycasting

Links