In the WWDC18 Platform State of the Union session, there was a cool example where a photo of a woman in front of one background was cut out, and placed onto a different background, using Vision/CoreML.
I want to do exactly this. I'm fine to do this on iOS12/beta only.
Does anyone have any resources on how to do this?
Oops, it was 54m:10-15s, and it's the Portrait Segmentation API. It is erroneously referred to in the WWDC 2018 State of the Union video by the presenter as the "People Segmentation API".