Handwritten Digit Recognition (MNIST + CNN)

This is a browser-based handwritten digit recognition project.
You can draw digits (0–9) on an HTML canvas, and a Convolutional Neural Network (CNN) trained on the MNIST dataset will predict the digit instantly using TensorFlow.js.

---

🚀 Features

• Draw digits directly in your browser.
• Real-time predictions with no server required.
• Pretrained CNN model (28×28 grayscale input).
• Simple and responsive user interface.

---

🛠️ Technologies Used

• HTML, CSS, JavaScript for the frontend.
• TensorFlow.js for running the model in the browser.
• MNIST Dataset for training.
• Python + TensorFlow/Keras for model training.