BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:An introduction to adversarial attacks and defences - Yingzhen Li 
 (University of Cambridge)
DTSTART:20171108T170000Z
DTEND:20171108T183000Z
UID:TALK95506@talks.cam.ac.uk
CONTACT:Adrià Garriga Alonso
DESCRIPTION:AI safety is not limited to RL settings. For example\, we can 
 use machine learning algorithms to design spam filters\, yet attackers can
  still “reverse-engineer” our defence to send us junk emails. Autonomo
 us driving systems based on computer vision techniques are also vulnerable
  to attacks\, for instance\, attackers can carefully apply a sticker to a 
 stop sign in order to fool the vision system of the car. In this talk I wi
 ll briefly discuss the mathematical framework of these attack techniques (
 specifically on image classifiers) and defence techniques against them.\n\
 nSlides available here: http://yingzhenli.net/home/pdf/attack_defence.pdf
LOCATION: Cambridge University Engineering Department\, CBL Seminar room B
 E4-38.  For directions see http://learning.eng.cam.ac.uk/Public/Directions
END:VEVENT
END:VCALENDAR