BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:From passive to interactive (multimodal) language learning - Angel
 iki Lazaridou\, University of Trento
DTSTART:20160415T110000Z
DTEND:20160415T120000Z
UID:TALK65224@talks.cam.ac.uk
CONTACT:Kris Cao
DESCRIPTION:The way humans learn the meaning of words is a fundamental que
 stion in\nmany different disciplines and\, from a computational perspectiv
 e\, an\nanswer to this question could lead to important advances in artifi
 cial\nintelligence. While the details of the learning process are still an
 \nopen question\, what we do know is that humans make use of a very rich\n
 perceptual input present in the communicative setups in which learning tak
 es place.\n\n\nIn this talk\, I will present our efforts on designing real
 istic multi-modal models  of human word learning. I will start by introduc
 ing  a model that assumes a purely passive learner existing in a non-commu
 nicative setup. I will then relax some of the learning assumptions and pre
 sent a model that assumes communicative episodes. Finally\, I will present
  our on-going work towards interactive learning between two agents.\n
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR