BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:NLIP Seminar Series
SUMMARY:Deep learning for automatically assessing the pron
 unciation of non-native English speakers - Kostas 
 Kyriakopoulos\, CUED\, University of Cambridge
DTSTART;TZID=Europe/London:20181026T120000
DTEND;TZID=Europe/London:20181026T130000
UID:TALK109621AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/109621
DESCRIPTION:In this talk\, I will present my work on automatic
 ally characterizing the pronunciation of non-nativ
 e English speakers based on spontaneous utterances
 . I will begin by defining what we mean by pronunc
 iation and the challenges presented by spontaneous
 \, as opposed to read-aloud\, speech. I will explo
 re two systems based on distances between phones: 
 a two-stage system using K-L divergence between Ga
 ussian models as an input to a DNN\, and an end-to
 -end system using Siamese LSTMs with attention ove
 r the hidden representations. It will be seen how 
 both systems can predict human-assigned grade\, as
  well as speaker L1 and country of origin\, and ho
 w their representations can be interpreted. Finall
 y\, I will discuss my approach to the detection of
  individual pronunciation errors and how it relate
 s to the auto-marking system.
LOCATION:FW26\, Computer Laboratory
CONTACT:Andrew Caines
END:VEVENT
END:VCALENDAR