BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Testing for Local Stationarity in Acoustic Signals: Parametric and
  Nonparametric Approaches - Daniel G. Rudoy\, School of Engineering and Ap
 plied Sciences\, Harvard University
DTSTART:20090914T101500Z
DTEND:20090914T111500Z
UID:TALK19311@talks.cam.ac.uk
CONTACT:Rachel Fogg
DESCRIPTION:This talk treats nonstationarity detection in the context of s
 peech and audio time series\, with broad application to stochastic process
 es exhibiting locally stationary behavior. Many such processes\, in partic
 ular information-carrying natural sound signals\, exhibit a degree of cont
 rolled nonstationarity that varies slowly over time. The talk first descri
 bes the basic concepts of these systems and their analysis via local Fouri
 er methods.\nParametric approaches appropriate for speech are then introdu
 ced by way of time-varying autoregressions\, along with nonparametric appr
 oaches based on time-localized power spectral density estimates\, along wi
 th an efficient offline bootstrap procedure based on the Wold representati
 on.  The talk includes asymptotic results as well as practical examples an
 d applications in speech forensics and audio waveform segmentation.\n\n
LOCATION:LR12\, Engineering\, Department of
END:VEVENT
END:VCALENDAR