This is a simple web-based digit recognition app using a trained Keras model and Flask for the backend. Users can draw a digit (0–9) in the browser, and the model will predict the number in real time. This is an introductory project for me to explore neural networks and deep learning for a common problem.
Model taken from here.
- Draw digits on a web canvas
- Classify handwritten digits (0–9) using a Convolutional Neural Network (CNN) trained on the MNIST dataset
- Real-time predictions using a Flask API
- Lightweight and easy to set up
- Python
- TensorFlow / Keras
- Flask
- HTML + CSS + JavaScript (frontend)
- Pillow (PIL) for image processing
- NumPy
git clone https://github.com/your-username/mnist-digit-classifier.git
cd mnist-digit-classifier
Open your terminal or command prompt and run:
pip install flask tensorflow pillow numpy