No. The computer vision model runs at the edge, on the player attached to the display, and no sensitive data leaves that device. The platform receives only anonymous audience attributes — such as the gender and age range of people passing the display — and uses them to choose which campaign to play.

Vision needs a webcam connected to the display, and is switched on per display with Enable Vision/AI (BETA) on the display's Enabled Features tab. The same tab controls whether the display records passer-by traffic and campaign engagement counts, and the Detection Preset that trades camera quality against CPU use. You can also run your own custom vision model at the edge; its complexity depends on the player hardware.

To set up gender or age targeting, see Target passers-by using computer vision at the edge.