In-Browser Face Recognition

This is a real-time face recognition web application built with React and the Face-API.js library for client-side facial analysis.

Implemented live webcam integration using the Web Media API to capture video streams and process them for face detection without server dependencies.

Integrated multiple pre-trained neural network models including TinyFaceDetector, face landmark detection, face recognition, and facial expression analysis for comprehensive face analysis.

Built a responsive canvas overlay system that renders real-time face detection boxes, 68-point facial landmarks, and emotion recognition directly on the video feed.

Developed asynchronous model loading with proper state management using React hooks to ensure all AI models are initialized before starting face detection.

Deployed as a lightweight, privacy-focused application on Vercel where all face processing occurs locally in the browser without sending data to external servers.