Handwritten Digit Recognition using CNN

How to Use the Demo

1. A Convolutional Neural Network (CNN) has been trained on the MNIST dataset, which contains thousands of 28x28 images of handwritten digits (0–9).
2. The model has learned to extract features from the input image (like strokes, edges, and shapes) and classify them into the correct digit.
3. Below, you will see a black drawing area. Use your mouse to draw any digit (0–9) in this box.
4. Once done, click on the Predict button below the box.
5. The model will process your input and display the predicted digit along with the probability.