BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Deciphering speech: a top-down approach to zero-resource speech re
 cognition - Peter Bell (University of Edinburgh)
DTSTART:20220601T140000Z
DTEND:20220601T150000Z
UID:TALK175241@talks.cam.ac.uk
CONTACT:Dr Jie Pu
DESCRIPTION:*Abstract*: In this talk I will present a method for training 
 an automatic speech recognition system for a language for which absolutely
  no transcribed training data is available – a so-called “zero resourc
 e” scenario.   Our approach is inspired by traditional methods of deciph
 erment used in code-breaking\, where strong prior knowledge of the languag
 e in question can be leveraged to crack a code with only limited quantitie
 s of cipher text.  When applying this technqiue to speech\, we use cross-l
 ingual knowledge transfer as a means of making the problem tractable\, and
  find that new languages can be “deciphered” with as little as 20 minu
 tes of audio\, and no phonetic knowledge of the language in question.\n\n*
 Bio*: Peter Bell is a Reader in Speech Technology in the School of Informa
 tics at the University of Edinburgh\, and a member of the Centre for Speec
 h Technology Research.  His research interests include all aspects of auto
 matic speech recognition development for low-resource languages.  He was p
 reviously a co-investigator on the IARPA MATERIAL project and is currently
  principal investigator on the Unmute project\, which aims to develop spok
 en language interfaces for traditionally marginalised language communities
 .
LOCATION:Zoom: https://eng-cam.zoom.us/j/81927138251?pwd=TVd3MXliV003dUdYV
 lFwU2NDWGpmdz09
END:VEVENT
END:VCALENDAR