Skip to content

Teach

Show your camera a few examples, then let it guess.

Camera is off

Turn it on to start teaching. Nothing is captured until you tap.

DINOv2-smallLoads on first use
  • 0/6
  • 0/6

Frames stay on this device. First use downloads about 42 MB.

How it works

Train turns your examples into image embeddings on this device, Classify embeds one frame, and your browser picks the closest label. Each frame is cropped to 256 px and embedded by DINOv2-small in this browser. Frames, labels and vectors stay in this tab's memory and are never uploaded or stored. Inspired by hunch.