it won't be though, it compromise privacy
When everything is voice-controlled, adversarial examples for speech recognition models may become a significant security vulnerability.
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
For those of us who wanto know what adversarial examples are, here is a starting point: https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45471.pdf …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
"activate administrator mode"
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I was wondering if human-inaudible "NN fooling" sounds are feasible
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This is related to speaker verification and spoofing detection.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
in speaker recognition it is called "spoofing" and is a field of study for some years now: http://www.sciencedirect.com/science/article/pii/S0167639314000788 …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@ceobillionaire what about people who can do accents, or change their pitch+tone easily/deliberately?Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
inb4
@OpenAI becomes more important than@DeepMindAI ;)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.