shaikfayaz2801-cmd/AI-Handwritten-Digit-Recognition

AI Handwritten Digit Recognition is a Streamlit web app that classifies user-drawn digits (0–9) using a Convolutional Neural Network trained on the MNIST dataset. It’s aimed at ML learners, developers and demo use — anyone who wants an interactive example of image classification. Key features: - Interactive drawing canvas (streamlit-drawable-canvas) - Real-time digit prediction with confidence score - Probability visualization across digits 0–9 - CNN model (TensorFlow/Keras) with ~98–99% MNIST accuracy - Dark-themed Streamlit UI Tech stack and structure: Python, TensorFlow/Keras, NumPy, Pillow; files include app.py, train_model.py, digit_model.h5 and requirements.txt. Usage: draw a digit, click “Predict”, the app preprocesses the image and returns the predicted digit and probabilities. Installation via pip install -r requirements.txt and run with streamlit.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.