BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Some new approaches to speech recognition and language modelling -
  Geoffrey Hinton (University of Toronto)
DTSTART:20110622T130000Z
DTEND:20110622T140000Z
UID:TALK31797@talks.cam.ac.uk
CONTACT:Kai Yu
DESCRIPTION:In the first part of the talk\, I will review some of the rese
 arch at the University of Toronto on the use of deep belief nets for speec
 h recognition. Then I will describe some very recent work that uses a rath
 er different approach to learn acoustic events that have precise times and
  amplitudes. Finally I will briefly describe a much better way of optimizi
 ng recurrent neural networks. When applied to a novel type of recurrent ne
 ural net\, this new optimizer learns extremely good models of character se
 quences. This is joint work with George Dahl\, Abdel-rahman Mohamed\, Navd
 eep Jaitly\, James Martens and Ilya Sutskever.
LOCATION:Cambridge University Engineering Department\, Lecture Room 2
END:VEVENT
END:VCALENDAR