Chess Engine with AI

A complete chess application with a graphical user interface and AI opponent using both traditional search techniques and neural network evaluation.

Project Status: Functional Prototype

Features

• Complete chess rules implementation (legal moves, castling, en passant, etc.)
• Graphical user interface with drag-and-drop piece movement
• AI opponent with adjustable difficulty levels
• Two AI evaluation options:
  • Traditional evaluation with alpha-beta pruning
  • Neural network evaluation (when trained model is available)
• Game state tracking (checkmate, stalemate, draws)
• Move history and undo functionality

Installation

1. Clone the repository
2. Create a virtual environment (optional but recommended):

   python -m venv .venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate
   

3. Install dependencies:

   pip install -r requirements.txt
   

Usage

Run the application:

python main.py

Controls:

• Click and drag to move pieces
• Use the difficulty dropdown to adjust AI strength
• Click "Undo Move" to take back your last move
• Press ESC to cancel a selected piece
• Press 'u' to undo a move

AI Implementation

The AI uses two evaluation methods:

1. Traditional Evaluation: A handcrafted evaluation function using material value, piece positions, and other chess heuristics with alpha-beta pruning search.

2. Neural Network Evaluation: A deep neural network that evaluates board positions. To use this feature, place a trained model file at assets/common/nn_model.pth.

Requirements

• Python 3.7+
• PyGame 2.5.2
• NumPy 1.26.3
• PyTorch 2.1.2
• python-chess 1.999

License

MIT

Acknowledgments

• Original inspiration by https://github.com/SebLague/Chess-Coding-Adventure
• Chess piece assets from https://github.com/lichess-org/lila