Home / Computer vision games you play with your body
How it works

Computer vision games you play with your body

Computer vision games use a camera to recognize movement and turn it into game input. In MoveToPlayGames, that can mean flapping to fly, swiping to slice fruit or holding a pose to clear a wall.

From camera to game control

After you give camera permission, pose tracking identifies body landmarks such as shoulders, wrists and hips. Each game looks for its own movement pattern: Flappy Wings looks for arm flaps, Panda Dash responds to a squat, and Shape Wall compares your pose with an outline.

Calibration helps the game check that it can see you and recognize the move. Good light, a stable camera and enough distance to fit your body in frame make a difference.

What happens to camera data?

Camera frames are processed locally on the device running pose tracking. With a directly connected camera, the game uses the detected landmarks on that device. In phone camera mode, normalized pose landmarks travel over an encrypted WebRTC data channel to the paired TV. The game does not send camera video to the TV or its pairing server. Read the full privacy policy, including the optional local Freeze Statue gallery.

Try different kinds of movement

Arms and hands

Upstream, Slice It and Punch Pads.

Whole-body timing

Log Jumper, Rocket Rope and Freeze Statue.

Browse all twelve games, then open the browser player to give one a try.