qalarc.com / projects / smart-glasses-face-tracker
Smart Glasses Face Tracker
Camera glasses stream to your home computer — it recognises faces, logs conversations, feeds you AI context via phone
Live browser demo of the smart glasses face tracking system. Uses MediaPipe FaceMesh (468 landmarks) and face-api.js for real-time face detection, recognition, and identity management. Auto-enrols new faces, tracks expressions, estimates age/gender/ethnicity, supports multiple simultaneous faces, and maintains a persistent face library. No server required — runs entirely in your browser. The project's core has since been extracted into the shared face-engine — the browser-first face capability layer that now also powers volkus.net's VKVerify — and it is part of the unified Face Tracking System platform.
Capabilities
Real-time face detection and recognition (30fps)
Auto-enrols and tracks multiple faces with stable IDs
Live expression analysis — 7 emotions with smoothed bars
Age, gender and ethnicity estimation per face
Persistent face library (localStorage, no server needed)
Face merge / dedup system with undo stack
Live 3D face mesh with Three.js (468 landmarks)
100% in-browser — webcam access only
Tags
Status: demo · First built: 2024-12-01 · Last updated: 2026-08-27