This project uses an AI pose estimation model to track human posture from your webcam and renders the results in a cross‑platform desktop UI.
- AI model:
yolo11m-pose.onnx, from Ultralytics’ YOLO11 models on Hugging Face. Link: https://huggingface.co/Ultralytics/YOLO11 - Rendering/UI: Built with Avalonia and Skia for fast, GPU‑accelerated drawing across platforms.
-
IDE tooling: It’s recommended to install the Avalonia extension for your IDE for better XAML tooling and previews.
-
Project templates: You can install the Avalonia .NET templates with:
dotnet new install Avalonia.Templates
- Captures webcam frames.
- Runs pose inference with
yolo11m-pose.onnx. - Renders keypoints and skeleton overlays using Skia within Avalonia views.
For the YOLO11 family and pose variants, see the Hugging Face page above. Ensure the ONNX model file (e.g., yolo11m-pose.onnx) is present in the app’s assets.
This project is licensed under the MIT License. See the LICENSE file for details.