AI & Computer Vision

AI-Based Proctoring System for Secure Video Interviews

Real-time webcam video analysis detecting suspicious activity with 90%+ accuracy.

View Source Code on GitHubLive Deployment Coming Soon

Project Impact & Architecture Overview

Developed an AI-based proctoring system for secure online video interviews using OpenCV, YOLO, and MediaPipe with instant alert logging in MongoDB.

Real-time monitoring through webcam video analysis

Detects suspicious activities: multiple faces, face absence, drowsiness, focus loss, and object usage

Generated instant alerts and stored logs in MongoDB

Achieved 90%+ accuracy with low latency

Python

FastAPI

OpenCV

YOLO

MediaPipe

MongoDB