Camera
Your webcam streams frames straight into the page. Nothing is recorded and nothing leaves this tab.
CT Learn · AI Coding · Level 4
A neural network finds 21 landmarks on your hand, sixty times a second — live in this tab. Draw in the air with a fingertip. Pinch to switch color. Make a fist to wipe the canvas.
All AI runs in your browser — nothing is uploaded.
hand_landmarker.task · float16 · ≈12 MB · runs on your GPU
Pick a mission
Same neural network, same 21 landmarks — eight different universes built on top.
How it works
Your webcam streams frames straight into the page. Nothing is recorded and nothing leaves this tab.
A 12 MB hand-landmark model runs right here — WebAssembly plus your GPU, about sixty predictions every second.
Each prediction is 21 (x, y, z) points per hand. Every gesture you'll see is just geometry on those points.
Warming up…
First run downloads a ≈12 MB model from the MediaPipe CDN — after that it's cached. The video itself never leaves your device.
We couldn't reach a camera on this machine. Hit Try again — or head back home; the landing page shows everything the tracker does.