| Job | Tool | Notes |
|---|---|---|
| Tracking / landmarks | MediaPipe FaceMesh | 468 points, 30fps on laptops, WASM |
| Model inference | ONNX Runtime Web | WASM; WebGL/WebGPU where available |
| Recognition | Embedding models + local store | Persistent identity library, all local |
| Demographics | Self-trained small models | We run 14 ("QalArc Folkus") — age bracket, gender, ethnicity, skin-tone zone |
| Heart rate (rPPG) | Signal filtering over face ROI | Green channel carries the pulse |
Every demo below opens your camera — then shows no video frames uploaded — after the one-time model/runtime loads, the network tab stays quiet during use. That's the honest on-device proof; demand it of any "private" face AI you evaluate.
Face tracking demo Heart-rate / rPPG demo The platform writeupModel downloads of a few MB happen once; after that, inference is free and offline. CPU-only machines handle tracking + small models; GPU helps for multi-face plus the demographic stack simultaneously.
Forensic-grade recognition or heavy multi-camera analytics still belong on server GPUs. The browser stack is for privacy-first, zero-install, single-user use — exactly what wearables and personal tools need.