View a demo video on YouTube here. Skip to 0:28 for said demo.
A C# Windows Store app using the Model-View-ViewModel design pattern that consumes data from the Kinect v2 sensor using Reactive Extensions (Rx).
The head vector scales depending upon the user's Z position, but this behaviour needs further tweaking.
I've separated my concerns to the best of my ability here. All of the Kinect data comes from a single Sensor class, and bitmap encoding only occurs in the view layer.