Handwritten Letter and Digits Identification

This is a project that uses TensorFlow and Keras to classify handwritten texts, identifying whether the input is a number or a letter. The model is trained to recognize both handwritten digits and letters, allowing it to distinguish between them with high accuracy.

Link to Trained Model -> https://github.com/JonatasMSS/handwritten_letter_digit

Technologies Used

• TensorFlow and Keras for building and training the neural network model.
• Python 3.11.9 for the development environment.
• NumPy for data manipulation.
• Streamlit for create a frontend interface

Installation Guide

Follow these steps to set up the project and run it on your local machine.

1. Clone the Repository

First, clone the repository to your local machine using the following command:

git clone https://github.com/JonatasMSS/handwritter_identifier.git

2. Set Up a Virtual Environment

It's recommended to use a virtual environment to manage dependencies. This ensures that the project uses the correct version of Python and all necessary libraries.

Create a Virtual Environment

Navigate to the project directory and create a virtual environment with Python 3.11.9:

python3.11 -m venv venv

If you don't have it, you can install using winget and after that use:

py -3.11 -m venv venv

Activate the Virtual Environment

• Linux/macOS:

  source venv/bin/activate

• Windows:

  venv\Scripts\activate

3. Install Dependencies

Once the virtual environment is activated, install the required dependencies listed in the requirements.txt file:

pip install -r requirements.txt

4. Running the Model

After installing the dependencies, you can run the model to classify handwritten letters and digits.

How to run

streamlit run main.py

The model will output whether the image contains a digit or a letter.

5. Deactivate the Virtual Environment

Once you're done working on the project, you can deactivate the virtual environment with the following command:

deactivate

Contributing

Contributions to the project are welcome. If you have any improvements, bug fixes, or new features, feel free to fork the repository, create a new branch, and submit a pull request.

License

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

---