The flip-side of the missing elephant in the room paper https://arxiv.org/pdf/1808.03305.pdf …pic.twitter.com/8SVRtvFcMU
Deep learning is having a rough year. Here’s the Fail of the Week. Can you see a person in this image? A TV? Neither can I. From https://arxiv.org/abs/1809.05966 .pic.twitter.com/iwywaTjOpq
-
-
-
Exactly
End of conversation
New conversation -
-
-
The contrast threshold is not high enough, I actually can see a human like figure very faint on that tile.
-
Do you think you see a human there? Or does top-down information make you think it’s at most a reflection, and probably not even that?
End of conversation
New conversation -
-
-
It's comical, but do you see this as a long-term, or insurmountable problem? Seems like the kind of fine-tuning one would/should expect at this stage. (I'm still a little amazed that my phone can recognize me, so perhaps I'm just easy to please).
-
I will have a whole book about this, roughly a year from now... short answer: it’s a long term problem to solve it correctly
- 1 more reply
New conversation -
-
-
as much as I generally agree w/you, if you look closely you can see what might be a person reflection where it says person (but it’s vague) & TV isn’t a horrible fail… there’s enough DLs to critique with really bad errors- maybe go easy on the “silly, but plausible"
-
But that’s the point. People use context and aren’t fooled.
- 3 more replies
New conversation -
-
-
If you play around with the contrast you can actually see a personpic.twitter.com/vG120wWFTD
-
The same way some people say they see Jesus in a slice of toast?
- 2 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.