Handwritten Digit Recognition – Web Application

An interactive web application that recognizes handwritten digits drawn by the user and returns real-time predictions from a trained machine learning model.

Live Demo:
https://handwritten-digit-recognition-vjhi.vercel.app/

---

What This App Does

• Accepts handwritten digit input via a drawing canvas
• Processes the input into a model-ready format
• Returns the predicted digit with class confidence scores

---

Stack

• React (frontend)
• Flask (backend API)
• Logistic Regression model trained on MNIST
• NumPy-based implementation (no scikit-learn)

---

Flow

Canvas Input → Image Processing → Backend Inference → Prediction Response

---

Run Locally

Backend

cd backend
pip install -r requirements.txt
python app.py

Frontend

npm install
npm start

---

Notes

This repository focuses on model integration and deployment. Detailed methodology and analysis are documented on the project webpage. → Project Webpage

---