Loading Patchies...

vision.face

vision

Detect face landmarks (478 points) or bounding boxes using MediaPipe

Inlets

video
Video input
MessageDescription
bangToggle enabled state
booleanSet enabled state directly
number0 disables, non-zero enables

Outlets

MessageDescription
Face landmarks (478 pts), optional blendshapes, or bounding boxes (detect mode)

Tags

mediapipefacelandmarkblendshapedetectionmlvisionmesh