Related Pages

FaceTrack

Use the camera to detect and track the position of a face

Overview

The FaceTrack widget uses the computer’s camera to find a face and
report its position, whether it’s smiling, and whether a face is
present at all — no training required, it works immediately on any
face in view.

Unlike a simple always-on camera widget, FaceTrack’s camera only turns
on when you want it to: an incoming value crossing a threshold
gates tracking on and off, the same pattern IfThen uses. Wire a
distance sensor, a button, or anything else to the input and the
camera only activates when that condition is met — useful for privacy
and for not running face detection continuously when nothing’s there
to detect. If you just want it running all the time, wire (or drag the
in-widget dial to) a value that stays above the threshold.

Main body

  • Camera Enable checkbox — Turns the camera on or off
  • Dial — shows the live input value that gates tracking on/off,
    and can be clicked and dragged to test the threshold without any
    hardware wired in at all.
  • Status dot — shows whether tracking is currently on, a face is
    currently detected, and whether it’s mid-wait on the threshold’s
    wait time (before starting or before latching off).
  • Status message — brief feedback for camera permission problems
    or detection errors.
  • Outlet readouts — live x, y, smile, and face (detected)
    values.

More panel

  • Threshold — the input value that turns tracking (and the
    camera) on.
  • Wait time true — how long the input must stay above threshold
    before tracking actually starts.
  • Wait time false (latch) — once tracking is on, how long the
    input must stay below threshold before tracking actually stops —
    a brief dip below threshold within this window doesn’t stop it.
  • Simulate (no camera) — bypasses the camera and threshold gate
    entirely, letting you manually set x, y, smile, and face detected with sliders/a checkbox to test everything downstream
    without a camera or a face in frame.
  • Camera preview — shown automatically once tracking starts, a
    live view of the camera with the tracked face points overlaid.

Outputs

  • x — horizontal center position of the detected face (0-1023).
  • y — vertical center position of the detected face (0-1023).
  • smile — how much the detected face is smiling, on a 0-1023
    scale.
  • face — whether a face is currently detected (0 = no, 1023 =
    yes).