No matter how much time is given, I feel the picture on the right will always trick me. Not sure how time plays a role.
. @techreview missed part of what is so interesting about this paper: humans are fooled only if the see images very briefly (a small fraction of a second); if people are given more time, they aren’t fooled (& neural nets consume vastly more energy doing the same task).https://twitter.com/techreview/status/969574859222257671 …
-
-
-
That picture is not representative, if you read the arXiv paper. Time details are supplied within it.
End of conversation
New conversation -
-
-
Great observation! And improvements given time suggests truly robust vision can benefit from *shhh* recurrent processing... But then how are all the deep nets working “so well” with purely feedforward connections..? (Hint: they’re not)
#aiThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I cringed at this comment "Most AI systems are loosely based on the human brain, after all." Do we know how the brain works to base AI on it?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Tech review has an unfortunate habit of coming up with dumb headlines like this.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This Tweet is unavailable.
-
This Tweet is unavailable.
-
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.