Live Camera Face Detection in .NET MAUI
Detecting faces, QR codes, or any visual pattern in a live camera feed requires getting frames into an ML pipeline efficiently. This post shows how to use DrawnUi.Maui.Camera and its SkiaCamera control to hook into the real-time video processing pipeline and run face landmark detection with MediaPipe — locally and consistently across iOS, Android, and Windows.
What you’ll learn
- How to enable the
SkiaCameraprocessing pipeline for AI/ML frame access withUseRealtimeVideoProcessing - How to set up MediaPipe local face landmark detection for cross-platform consistency
- How to send live video frames to both local ML engines and remote vision API endpoints
- How to draw face detection overlays and make face masks stick to moving heads in real-time
- The difference between local ML detection (latency, offline support) and API-based detection (flexibility, model variety)
- Architecture considerations for building detection features into .NET MAUI apps