StreetFighter-GameBot

• Trained an RL agent to play Street Fighter in real time using reward-based learning • Used computer vision for game-state perception • Designed a custom reward function that incentivised aggressive, adaptive combat • Agent successfully learned non-trivial attack sequences

GameBot: Machine Learning Bot for Street Fighter II Turbo

This project implements a machine learning-based bot that plays Street Fighter II Turbo by analyzing real-time game state data using PyTorch.

How to Run

Prerequisites

• Python 3.6 or higher
• Install dependencies:

  pip install -r requirements.txt