Load MediaPipe selfie segmentation engine in TouchDesigner and build an alpha mask adapter. Optionally invert, feather, or output person on transparent. Wire into any matte-consuming tool.
Renders a connected stick-figure skeleton from full-body pose tracking. Supports synthetic, MediaPipe, OSC, or existing pose data with adjustable line color, width, and camera distance.
Run any model from the Replicate catalog by its owner/name identifier. Supports all categories including embeddings, segmentation, depth estimation, and more.
Builds a depth-driven GPU point cloud field using a depth or mask TOP, with displace, emit, or both scatter modes for displacement and particle effects.
Load the MediaPipe body tracking engine into your TouchDesigner project, start the timeline, and build a live 33-landmark pose skeleton from webcam video. Guides installation if the engine is missing.
Integrates Square sandbox catalog and orders management with voice commands (via ElevenLabs) and hand gesture controls (via MediaPipe) through an interactive seller dashboard in Goose Desktop.
Extracts deterministic facial features from photos locally using MediaPipe, returning structured JSON with measurements like face shape ratios and symmetry.
Configures MediaPipe hand tracking from a webcam in TouchDesigner. Produces a CHOP with 21 hand landmarks per hand for gesture detection or skeleton creation.
Build a body-reactive visual with glowing marks that follow 33 pose landmarks, rendered to a Null TOP. Supports points, glow, or trails styles; default synthetic source works without camera or plugin.
Derives reactive scalar channels from MediaPipe pose landmarks for body-driven parameters. Supports metrics like position, distance, angle, velocity, and openness with optional direct parameter bindings.
Creates a component that transforms hand landmarks into debounced gesture channels for controlling palm holograms, lasers, audio, and other hand-reactive visuals in TouchDesigner.
Set up full-body pose tracking with 33 MediaPipe landmarks and ready-to-bind keypoints. Use synthetic mode for preview without camera, or switch to mediapipe, OSC, or existing CHOP for real performer input.
Enables real-time face landmark tracking from a webcam by loading the MediaPipe engine, starting the timeline, and outputting a 468-sample CHOP for direct use in visualization or binding.
Build a hand gesture control network for Ableton Live using MediaPipe, outputting pinch and wrist roll maps with a skeleton overlay for performance mapping.
Creates an interactive bubble simulation over a live camera: an open palm emits bubbles, tracked body and hand landmarks act as colliders to bat or lift them, with a visible body contour for clear interaction.