News for hackers. Stuff that matters.

A comment on Psychedelic stickers that interfere with AI image recognition. Parent comment

Re:Psychedelic stickers that interfere with AI image recogni... by zitterbewegung on Sunday January 07, 2018 @10:04PM (#16094346)
Hardening production ML systems are going to be fun. I think exploiting ML will be the new kid on the block just like how we were introduced to XSS exploits. See https://github.com/cchio/deep-pwning
[ Parent | on HN ]
Re:Psychedelic stickers that interfere with AI image recogni... by mtgx on Sunday January 07, 2018 @10:37PM (#16094468)
Indeed. It's already starting to happen:

https://www.csail.mit.edu/news/fooling-googles-image-recogni...

[ Parent | on HN ]
Re:Psychedelic stickers that interfere with AI image recogni... by username223 on Sunday January 07, 2018 @10:52PM (#16094532)
Trying to "harden" million-parameter models trained on a relatively small number of relevant examples will be a nightmare to make web security look easy.

PS -- NIPS is in Long Beach now? What a shame.

[ Parent | on HN ]
Re:Psychedelic stickers that interfere with AI image recogni... by robotresearcher on Sunday January 07, 2018 @11:09PM (#16094581)
It left Vancouver a few years ago and has been moving around. It's getting too big for many cities now.
[ Parent | on HN ]
Re:Psychedelic stickers that interfere with AI image recogni... by username223 on Sunday January 07, 2018 @11:38PM (#16094701)
Too bad... Vancouver/Whistler was an awesome place to have a conference, even if the weather was rainy and the skiing so-so. A NIPS too big to be hosted by anything but a mega-city sounds depressing.
[ Parent | on HN ]