BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Speech Recognition:  What’s Left? - Dr Michael Picheny
DTSTART:20191112T120000Z
DTEND:20191112T130000Z
UID:TALK133159@talks.cam.ac.uk
CONTACT:Lina Zvaginyte-Bagociene (div-f)
DESCRIPTION:Recent speech recognition advances on the SWITCHBOARD corpus s
 uggest that because of recent advances in Deep Learning\, we now achieve W
 ord Error Rates comparable to human listeners. Does this mean the speech r
 ecognition problem is solved and the community can move on to a different 
 set of problems? In this talk\, we examine speech recognition issues that 
 still plague the community and compare and contrast them to what is known 
 about human perception. We specifically highlight issues in accented speec
 h\, noisy/reverberant speech\, speaking style\, rapid adaptation to new do
 mains\, and multilingual speech recognition. We try to demonstrate that co
 mpared to human perception\, there is still much room for improvement\, so
  significant work in speech recognition research is still required from th
 e community.
LOCATION:Department of Engineering - LT1
END:VEVENT
END:VCALENDAR