../ machine learning

pose

Detect key points of the human body using MediaPipe.

  • closest-landmark: for a single pose, find the landmark closest to a target landmark.
  • shoulders-basic: demonstrates basic calculation from points.
  • hands-arms: demonstrates basic calculation using hand and arm landmarks.
  • head: Simple calculation based on a few points.
  • between-basic: access points from a spatial basis.
  • shoulders: advanced version, uses data to influence a Thing.
  • between-advanced: works with distance between bodies. Uses the DOM and Things model.
  • landmark: Shows historical points per landmark
  • starter-canvas: basic starting point to work with pose detction data and the canvas.
  • starter-dom: basic starting point to work with pose detction data and the DOM.
  • sender: generate and emit body pose data.

hand

Detect key points of the hand using MediaPipe.

  • pinch: Calculate a pinch gesture.
    Interpolation
  • techniques: Demonstrates a few inferences from data
    Points.distance, Numbers.average, Numbers.scaler
  • palm: Control things using area of palm
    Things
  • sender: generate and emit hand pose data.

face

Detect a few key points of the face using MediaPipe.

  • demo: compute angles of face movement
  • sender: generate and emit face pose data.

objects

Detect a small number of objects using MediaPipe.