Skip to content

Latest commit

 

History

History
23 lines (16 loc) · 1.2 KB

README.md

File metadata and controls

23 lines (16 loc) · 1.2 KB

Digit Recognition Web Application

Click here to try it out yourself: Deployed Project Link

React.js (frontend) and Flask (backend) web application that incorporates a neural network to identify user-drawn digits.

I am incorporating the trained PyTorch neural network model from my MNIST Neural Network project for identification as a part of the Flask backend.

Description

  • The user draws the digit and hits submit, which sends the image's URL to the Flask backend using a POST request
  • The Flask server then preprocesses the image and converts it into a tensor
  • Afterward, the tensor is given to the pre-trained neural network model for identification, which results in a predicted digit
  • The digit is then sent back to the React.js frontend where it will be displayed on the table
  • The Flask API backend is hosted on an AWS EC2 Instance and the React.js frontend is hosted on Netlify
    • The AWS server uses Cloudfront to redirect HTTPS requests to the HTTP Flask server, since requests from Netlify are HTTPS

Demo

Demo