If you want to be the one-stop, low-cost leader in motion capture, you need to integrate facial recognition and capture.

I'm happy to pimp you guys to my colleagues, who can make use of this, but it's unrealistic that we have to dump our existing camera system and software and jump into an entirely different suite if we are interested in moving faces.

There is software out there for Kinect V2 and Iphone X, facial mocap. So I know the Azure can handle it. It doesn't have to be captured simultaneously with body motion. You could require a narrow field, dedicated, neck-up view. But we need to be able to integrate these.

Thanks.

This is simple test of what Azure Kinect can do as far as native facial motion capture.

https://hardluckhank.com/wp-content/upl ... ovie-3.mp4

And these are the 52 blendshapes that Apple tracks with its TrueDepth camera on IphoneX+.
https://developer.apple.com/documentation/arkit/arfaceanchor/blendshapelocation

I think the Azure Kinect can produce some decent facial mocap. With 2+ stitched together you could cover a face easily and error correct.

REQUEST.

Blender now has the ability to track your face movements with a webcam. Fast frame rates are better. You just need to put dot targets, either pen or sticky dots on your face and you can capture face movements that can be transfered to a character. I've seen it working but not being a blender user myself I found the transfer to a character a bit complicated because I don't know the program well enough. So if blender can do it for free, I'm sure ipi could come up with a smoother and easier system to work with. Which would be an incredible bonus to us animaters