BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Dialogue management personalisation for users with speech disorder
 s[CANCELLED] - Iñigo Casanueva
DTSTART:20160602T110000Z
DTEND:20160602T120000Z
UID:TALK66235@talks.cam.ac.uk
CONTACT:Anton Ragni
DESCRIPTION:Voice controlled environmental control interfaces can be a lif
 e changing technology for users with disabilities. However\, often these u
 sers suffer from speech disorders (e.g. dysarthria)\, making  ASR very cha
 llenging. Acoustic model adaptation can improve the performance of the ASR
 \, but the error rate will still be high for severe dysarthric speakers. P
 OMDP-based dialogue management can improve the performance of these interf
 aces due to its robustness against high ASR error rates and its ability to
  find the optimal dialogue policy in each environment (e.g. the optimal po
 licy depending on the dysarthria severity of the speaker or on the amount 
 of acoustic data used to adapt the ASR). However\, very little research ha
 s been done so far in dialogue model adaptation to unique users interactin
 g with a system over a long period of time. This talk explains how statist
 ical dialogue management can be applied to an environmental control interf
 ace such as homeService. It also explores different methods to adapt Gauss
 ian process-based dialogue policies and RNN-based dialogue state trackers 
 to unique users\, and methods to adapt these models online as more data fr
 om the user becomes available.
LOCATION:Department of Engineering - LR5
END:VEVENT
END:VCALENDAR