“Universal adversarial perturbations” seems most dramatic ML result in years; if so, not getting deserved attention https://arxiv.org/pdf/1610.08401v1.pdf …
-
-
Disclaimers: I don't follow this field closely, may be obviously wrong, and have done no experiments to test the theory. (Might like to!)
-
Thanks to Jeff Shrager (https://explorecourses.stanford.edu/instructor/jshrager …) for drawing the paper to my attention.
- 2 more replies
New conversation -
-
-
“Measuring the tendency of CNNs to Learn Surface Statistical Regularities“ https://arxiv.org/abs/1711.11561
-
Thanks!! This is nice work (and supports my theory about why DL image classifiers work…) I’m trying to resist writing a long tweetstorm about it, but I’m not sure I’m going to succeed!
End of conversation
New conversation -
-
-
millions of years of evolution under adversarial circumstances, lighting, movement, camouflage; very different learning results
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
That CNNs do not well capture higher order shape information and focus mostly on texture and color distr is well studied.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.