Spaces:
Sleeping
Sleeping
metadata
title: DigitVision
emoji: 🔢
colorFrom: blue
colorTo: purple
sdk: docker
sdk_version: '3.11'
app_file: app.py
pinned: false
DigitVision
Handwritten Digit Recognition with CNN
A full-stack web application for recognizing handwritten digits using a trained CNN model with 98% accuracy.
Features
- Upload digit images for prediction
- Draw digits on canvas with pencil/eraser tools
- Real-time predictions with confidence scores
- Modern neon-themed UI with particle animations
- Sign in/Sign up functionality
Tech Stack
- Backend: Flask, TensorFlow, Keras
- Frontend: HTML, CSS, JavaScript, Bootstrap
- Model: CNN trained on MNIST dataset (98% accuracy)
Usage
- Visit the app
- Sign in or sign up
- Upload a digit image or draw one
- Get instant predictions!