Chess Piece Classification using CNN

HOW TO RUN PROJECT

1. Download the Jupyter notebook:

   wget https://github.com/rezapace/Machine-Learning-chess-classification/releases/download/1.0/klasifikasi_catur.ipynb

2. Open Google Colab: https://colab.research.google.com/#create=true

3. Upload the downloaded klasifikasi_catur.ipynb file to Google Colab.

4. Run the notebook:

   • Execute each cell in order
   • Follow the instructions provided in the notebook comments

Note: Make sure you have a Google account to use Google Colab. If you encounter any issues, please refer to the project repository for troubleshooting or to report problems.

Description

This project implements a Convolutional Neural Network (CNN) to classify chess pieces. It uses TensorFlow and Keras to build and train a model that can identify five different chess pieces: king, queen, rook, bishop, and pawn.

Features

• Image classification of chess pieces
• Utilizes Convolutional Neural Networks (CNN)
• Data augmentation for improved model performance
• Easy-to-use Google Colab implementation

Requirements

• TensorFlow
• Keras
• scikit-learn
• Google Colab (for easy execution)

How to Run

1. Download the Jupyter notebook file: klasifikasi_catur.ipynb

2. Open Google Colab

3. Upload the downloaded .ipynb file to Google Colab

4. Run the following commands in a Colab cell to clone the repository and set up the environment:

   !apt install git
   !git clone https://github.com/rezapace/Machine-Learning-chess-classification

5. Execute the cells in order:

   • Import required libraries
   • Prepare the dataset
   • Set up data augmentation and loading
   • Build the CNN model
   • Compile the model
   • Train the model
   • Evaluate the model
   • Save the model (optional)

6. The final cell will display the model's accuracy on the validation set.

Model Architecture

The CNN model consists of:

• 3 Convolutional layers with ReLU activation
• 3 MaxPooling layers
• Flatten layer
• Dense layer with ReLU activation
• Dropout layer (0.5) for regularization
• Output layer with Softmax activation for 5 classes

Results

The model achieves an accuracy of approximately 42.40% on the validation set after 10 epochs of training.

Future Improvements

• Increase the dataset size
• Experiment with different model architectures
• Implement transfer learning using pre-trained models
• Fine-tune hyperparameters for better performance

Contributing

Contributions to improve the model or extend the project are welcome. Please feel free to fork the repository and submit pull requests.

License

This project is open-source and available under the MIT License.

---

For more information or to report issues, please visit the project repository.