Facial Expression Identification in .NET MAUI on iOS | Community Feed

ARKit can identify over 50 facial expressions, often several at once. David Britch builds on face tracking to read a user’s expression in a .NET MAUI iOS app and react to it by recoloring the overlaid face mesh.

What you’ll learn

Rendering the face mesh — creating an ARSCNFaceGeometry when an ARFaceAnchor is detected
Reading blend shapes — interpreting floating-point values (0 to 1) that represent the presence of each expression
Detecting expressions — thresholding blend shapes for blinks, wide eyes, frowns, smiles, raised brows, tongue-out, and more
Reacting in real time — updating the mesh geometry and color in DidUpdateNode as expressions change
Where this leads — the foundation for gaze tracking and emotion recognition

Read the full post for the expression-detection code, and note that ARKit requires a physical iPhone.