so far: looks like the neural net has only a fuzzy idea of what it's actually seeing & relies heavily on probability.
-
-
Show this thread
-
this suggests a hilariously surrealist method of fooling image recognition-based security.
Show this thread -
This is the neural net I've been using, by the way: https://github.com/karpathy/neuraltalk2 … And you can play with Microsoft's version here:https://azure.microsoft.com/en-us/services/cognitive-services/computer-vision/ …
Show this thread -
I summarized the experiments in a blog post here. Thanks for all the help, everyone!http://aiweirdness.com/post/171451900302/do-neural-nets-dream-of-electric-sheep …
Show this thread -
so obvious here but think of the mistakes algorithms can quietly make w medical & satellite photos. Expert review is crucial.
Show this thread
End of conversation
New conversation -
-
-
How about photographs of those concrete sheep statues that seemed to be in vogue in city centers a couple of years back?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I've seen some weird Catan configurations.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
"a white and black sheep standing on a sidewalk"
- Show replies
New conversation -
-
"A white cat is sitting on a windowsill"
- Show 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.