Jul 30, 2017 · 1 min read
Agree with a lot of things here. Especially point 3 and point 4 — deep learning models won’t recognise that it’s a reflection because they haven’t spent three years in a world which causes them to learn the difference. I’m sure a deep learning model, given the same amount of sensory data as humans could distinguish between a real image and reflection easily. I think you’re hitting the nail on the head here.
Thanks for the shoutout too :)
