Send body and hand tracking data from your webcam to any software that speaks OSC (like Wekinator, Max/MSP, Pure Data, TouchDesigner...).
Both scripts send to /wek/inputs at 127.0.0.1:9000 — ready to plug into Wekinator directly.
Step 1 — Open File Explorer and go to C:\
Open any folder window, then click on the address bar at the top, type C:\ and press Enter.
Step 2 — Open a terminal here
Right-click on an empty area of the folder (not on a file) and select "Open in Terminal".
If you don't see that option, hold Shift while right-clicking.
Step 3 — Run these four commands, one by one
git clone https://github.com/cbicari/c-lab-scripts
Set-ExecutionPolicy Bypass -Scope Process
cd c-lab-scripts
.\install.ps1That's it. The install script takes care of everything: Python 3.12, the virtual environment, all dependencies, and two shortcuts on your desktop — Hand Tracking and Body Tracking.
Step 4 — Use the desktop shortcuts
Just double-click a shortcut to launch. Launching one will automatically close the other. Press Q or close the window to stop.
Why
C:\? Windows usernames with accented characters (e.g.Étudiant) break MediaPipe. Installing atC:\avoids this entirely.
Step 1 — Open a terminal in the project folder
Step 2 — Run
git clone https://github.com/cbicari/c-lab-scripts
cd c-lab-scripts
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txtStep 3 — Launch a script
python hand_recognition.pypython mediapipe_body.pyPress Q or close the window to stop.
| Setting | Value |
|---|---|
| OSC input port | 9000 |
| OSC address | /wek/inputs |
| Number of inputs | 63 (hand) or 66 (body) |
Camera not opening — Try detection_context(dev_id=1) in the script to switch camera.
ModuleNotFoundError — Your terminal is not using the venv. On Linux/Mac run source venv/bin/activate first. On Windows, use the desktop shortcuts instead.
Slow or laggy — MediaPipe runs on CPU. Close other heavy applications.