Digging into the developer documentation for Quest 3 I see there's an API for Scene Understanding that utilizes AI/ML for recondition of common objects like ceiling, wall, door, desk, lamp, etc... but what I do not see is the means to recognize objects. For example if I wanted to create a MR training simulation for Quest 3 that involves object condition of a custom computer control panel and then provides augmented dialogs (see picture below). I don't see any means to provide pictures of custom objects for the recondition to occur.
Compatibility with developer.vuforia.com could provide this, but it seems Meta will not allow imagery to be shared with the app running on the device making object and image recondition style AR/MR apps impossible. Is this correct?
