BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Generative Adversarial Networks - Max Campman\, Churchill College
DTSTART:20190130T193000Z
DTEND:20190130T200000Z
UID:TALK119404@talks.cam.ac.uk
CONTACT:Matthew Ireland
DESCRIPTION:Generative Adversarial Networks (GANs) consist of a pair of ne
 ural \nnetworks: a generator and a discriminator. The two compete in a min
 imax \ngame where the generator aims to create data which the discriminato
 r is \nunable to distinguish from a genuine dataset. This game enables the
  \ncreation of deep generative models\, a research topic which previously 
 \nhad little success.\n\nI will be summarising the original game as propos
 ed by Goodfellow et al. \nand will be looking at how such a game may be mo
 dified to fit more \nspecific criteria with a particular focus on an image
 -to-image \ntranslation system called CycleGAN.
LOCATION:Wolfson Hall\, Churchill College
END:VEVENT
END:VCALENDAR