Just as we can get ARKit to detect pre-defined 2d images, believe it or not, we can get it to detect pre-defined real world 3d objects.
I believe this is done by effectively 'scanning' a 3d objects feature points and using them as a point of reference for comparison later on when trying to detect the 3d object again.
I'll know more once I get to writing the code for this lesson.
No video yet
No code yet
Next Step : Vision and Core ML
After you have mastered this you should try Vision and Core ML