Tech Stack - Python, Tensorflow, OpenCV, Flask, HTML5, CSS3, JavaScript
Algorithm - Convolutional Neural Networks

• The CNN based classifier is capable of recognizing Hand written digits.
• The Model was trained on MNIST Dataset.
• Using a custom architecture an accuracy of around 98% was achived when tested on 10000 test images.
• The project was deployed on a Flask web server
• The Front End was designed using HTML, CSS, JS.
• HTML canvas was used to give the user a functionality to draw in real time.