ARKit Face Recognition Integration

[dokun1]

With the iPhone X, you can capture a very granular 3d map of someones face from the selfie camera. It should be possible to stream this map in realtime from the camera if a setting for this is on.

[dokun1]

For more documentation, check out: https://developer.apple.com/documentation/arkit/creating_face_based_ar_experiences

[dokun1]

We can update this API to use Vision for this API. I envision an API which allows you to switch between CoreImage and Vision (when available).

[amfang]

I'm trying to integrate ARkit and Lumina together, but it's not working, any idea to this??

[dokun1]

@amfang I think the main issue is that ARKit provides an entirely integrated camera. If we were to use the ARKit view controller methodology, Lumina would lose the ability to control the camera controls itself, or at least that's what I am thinking.

This issue might be a non-starter.

[amfang]

@dokun1 I checked Apple document for Vision at coreML for object detect with AR for Machine-Learning Image Analysis, although can't streamed, but can reach the same effective like Lumina.