[IIAOPSW]

>plus as-of-yet unknown human vision to synthesize depth info correctly for arbitrary scenes

I have an adversarial example.

https://wallpapercave.com/wp/InAcQKW.jpg