Instructions for developers who want to run the plugin from the source code.
- Python 3.11
pip install -r requirements.txt
note: The plugin use the method vts.vts_request.requestSetMultiParameterValue
from the library pyvts.
This method is not included in the latest released version 0.3.2. You will need it to run the plugin, which can be obtained from the current repository. Required method: https://github.com/Genteki/pyvts/blob/main/pyvts/vts_request.py l.246
- Open Vtube Studio
- Start the plugin
python app.py
pyinstaller ./app.py -n VTS_Fullbody_Tracking-0.1.10 --add-data='models/*:models' --add-data="venv/Lib/site-packages/mediapipe/modules/pose_landmark/*:mediapipe/modules/pose_landmark" --add-data="venv/Lib/site-packages/mediapipe/modules/pose_detection/*:mediapipe/modules/pose_detection" --add-data='*.png:.' -F -w
mediapipe documentation -Available Pose Landmarker models
mediapipe documentation - landmarker python
Blog Research Google - On-device, Real-time Body Pose Tracking with MediaPipe BlazePose
VTube Studio API: https://github.com/DenchiSoft/VTubeStudio
nizima api: https://live2d-garage.github.io/
nizima Live API doc: https://github.com/Live2D/nizimaLIVEPluginAPI/tree/develop
mediapipe: https://pypi.org/project/mediapipe/
pyvts: https://genteki.github.io/pyvts/
pyinstaller: https://pyinstaller.org/en/stable/
pynizima: https://github.com/jellydreams/pyNizima