Experience your past and future self using real-time AI face tracking.
How AgeSwitch AI Works
This experiment runs entirely in your browser. No images or video data are ever uploaded to a
server. It utilizes vanilla JavaScript, the HTML5 Canvas API, and Google's MediaPipe machine
learning framework.
The Technology Behind the Effects
- Biometric Face Tracking: MediaPipe's Face Mesh detects 478 unique facial landmarks in
real-time, adapting to your face shape, scale, and rotation at 60 frames per second.
- The "Youth" Algorithm: Simulates neoteny (youthful features) using digital soft-focus
techniques. The engine duplicates your camera feed, applies a real-time blur, and merges it back using a
screen blend mode for a radiant skin glow. It applies dynamic digital blush and
geometrically enlarges the eyes.
- The "Aging" Algorithm: Employs texture anchoring and procedural generation. Age spots
are generated mathematically and anchored specifically to the bridge of your nose, ensuring they rotate
and tilt naturally with your head movements.
- 3D Wrinkle Emulation: Fine lines and wrinkles are mapped to your forehead, nasolabial
folds, and crow's feet. Using an "emboss" rendering technique (offsetting light and shadow curves), the
2D canvas creates a convincing illusion of 3D facial depth.
Privacy First
Because all processing happens via WebAssembly and JavaScript locally on your device's CPU/GPU, your webcam
feed remains strictly private.