BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Towards understanding the performance of individuals within automa
 tic speaker recognition systems - Vincent Hughes\, University of York
DTSTART:20240624T110000Z
DTEND:20240624T120000Z
UID:TALK217933@talks.cam.ac.uk
CONTACT:Simon Webster McKnight
DESCRIPTION:Automatic speaker recognition (ASpR) systems are widely used i
 n commercial settings for the purposes of personalisation and security\, a
 nd are now increasingly being used to evaluate forensic evidence presented
  in court. Despite improvements in overall performance\, even with forensi
 cally realistic materials (Morrison and Enzinger 2019)\, very little is kn
 own about why certain voices would be easy or difficult for systems to rec
 ognise. \n\nIn this talk\, I will introduce a current ESRC-funded project 
 (Person-specific ASR) which attempts to better understand and explain indi
 vidual variability in ASpR performance\, with a particular focus on forens
 ic applications. I will present data from a series of studies using both a
  small-scale\, highly controlled dataset of recordings containing extreme 
 vocal variation\, as well as a large-scale\, forensically realistic databa
 se provided to us by the UK Government. Finally\, I will discuss our initi
 al attempts to handle problematic speakers for ASpR systems\, through cond
 ition adaptation and tailored calibration.\n
LOCATION:Zoom only: https://cam-ac-uk.zoom.us/j/89860790207?pwd=sbErkKGhMw
 BJhyJcmlJ1jcQml5oTd4.1
END:VEVENT
END:VCALENDAR