Handwritten Digit Recognition using CNNs

Overview

This project is a Handwritten Digit Recognition system built using Convolutional Neural Networks (CNNs) with TensorFlow/Keras. It is trained on the MNIST dataset, which contains 60,000 training and 10,000 testing grayscale images of handwritten digits (0-9). The model achieves high accuracy in recognizing digits from user input.

Features

• Trained on the MNIST dataset.
• Uses CNNs for high accuracy.
• Supports real-time digit recognition from images.
• Interactive UI (if applicable) for testing.
• Can be extended for custom datasets.

Model Architecture

• Conv2D layers for feature extraction
• MaxPooling2D for dimensionality reduction
• Flatten to convert 2D features into a 1D array
• Dense (Fully Connected Layers) for classification
• Softmax Activation for output prediction

Results

• Training Accuracy: ~99%
• Test Accuracy: ~98%