Fyrecracker.
Stopped:
.This project is currently STOPPED.
Fyrecracker is a project that intends to blend gameplay and deep learning artificial intelligence. Be the last of your friends standing by exploding them out of the playable area, or outlast AI agents trained through reinforcement learning algorithms. This project is created using Unreal Engine 5, and strives to produce a learning environment for AI entities.
⚪ Check out the GitHub repository. Click here.
⚪ Fyrecracker is based on a Minecraft minigame of the same name. Check out the original project here.
⚪ Unreal Engine 5 is a next-generation game engine developed by Epic Games. Check out Unreal Engine here.
▱▰▱ This project will include deep learning AI agents. ▰▱▰
- 1. Agents will be trained with reinforcement learning.
- 2. Hundreds of 2-minute simulations on small specialized maps will undergo, and the last remaining AI will be given points based on the efficiency of eliminating the other agents.
- 3. The agent will be fully trained through a neural network. This includes movement, firing, and strategy.
- 4. The goal of providing small-scale rapid testing is to train the AI to be aggressive, as well as reinforce positive survival strategies.
- 5. Check back with this project in the future. There will be a video explanation of the entire process.
