SuperVision AI — Real-time student engagement monitoring using Face Mesh & WebSockets with separate Student & Teacher dashboards.

Tagline -

An AI-powered classroom supervision system that detects student engagement in real time and visualizes it live for instructors.

Overview

SuperVision AI is a real-time AI-based classroom monitoring system designed to analyze student engagement using facial landmark detection and stream live insights to a teacher dashboard.

The system uses MediaPipe Face Mesh to detect facial expressions, FastAPI WebSockets for real-time communication, and a React + Vite frontend to display live engagement states and trends.

Key Features

Live Face Detection using MediaPipe Face Mesh

Engagement Classification (Focused / Neutral / Confused)

Real-time WebSocket Communication (Student → Teacher)

Teacher Dashboard with live status & engagement timeline

Student Portal with face mesh overlay & feedback

Low-latency, event-driven architecture

Tech Stack

Frontend

React.js (Vite)

MediaPipe Face Mesh

Chart.js

WebSockets

Backend

FastAPI

WebSockets (Starlette)

Python

Architecture

Student (Camera + Face Mesh) ↓ Engagement State (WebSocket) ↓ FastAPI Backend ↓ Teacher Dashboard (Live Updates + Chart)

Use Cases

Online classes & virtual classrooms

Proctoring & attention monitoring

EdTech platforms

AI-based behavioral analytics projects

Why This Project?

This project demonstrates:

Real-time AI inference in the browser

WebSocket-based system design

Clean separation of Student & Teacher roles

Practical application of computer vision in education

It was built as a personal learning project to explore real-time AI systems and full-stack integration.

Author

Anchal Gupta

Aspiring Data Scientist | AI & Full-Stack Enthusiast