Handwriting Recognition ✍🏻💻

This project features a handwriting recognition system powered by a Convolutional Neural Network (CNN), built using TensorFlow and Keras. It is designed to accurately recognize handwritten digits, leveraging data from the MNIST dataset.

📊 Model performance

Training and validation accuracy

The graph below shows the training and validation accuracy of the model :

⚙️ Features

• Train a CNN model on the MNIST dataset.

• Save and load the trained model.

• GUI for loading images and drawing digits on a canvas.

• Real-time prediction of handwritten digits with confidence scores.

• Visualization of training history and predictions.

  

🛠️ Installation

1. Clone the repository:

   git clone https://github.com/soroqn1/Digit-Recognition
   cd Digit-Recognition

2. Install the required dependencies:

   pip install -r requirements.txt

🔍 Usage

Training the Model

1. Run the general.py script to train the model:

   python general.py

2. The trained model will be saved as models/handwriting_recognition_model.h5.

Running the GUI

1. Run the gui.py script to start the GUI application:

   python gui.py

2. Use the GUI to load an image or draw a digit on the canvas for prediction.

🗂️ Files

• general.py: Script for training the CNN model on the MNIST dataset.
• gui.py: Script for the GUI application to load images and draw digits for prediction.
• requirements.txt: List of required Python packages.

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements

• The MNIST dataset is provided by Yann LeCun and Corinna Cortes.
• TensorFlow and Keras are open-source libraries developed by the TensorFlow team.