BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Monaural Acoustical Scene Analysis through Harmonic-Temporal Clust
 ering of the Power Spectrum - Jonathan Le Roux\, Ecole Normale Superieure\
 , Paris\, France and University of Tokyo
DTSTART:20071112T150000Z
DTEND:20071112T163000Z
UID:TALK9100@talks.cam.ac.uk
CONTACT:Taylan Cemgil
DESCRIPTION:The design of effective algorithms for single-channel analysis
  of complex and varied acoustical scenes is a very important and challengi
 ng problem. We present here a framework called Harmonic-Temporal Clusterin
 g (HTC) which relies on the description of the power spectrum as a combina
 tion of constrained Gaussian Mixture Models. The parameters of the models 
 are simultaneously estimated through a global fitting to the observed powe
 r spectrum in the time-frequency domain. The optimal solution can be used 
 both for F0 estimation in various noisy environments as well as in multipl
 e speaker situations\, and to perform single channel speech enhancement\, 
 background retrieval and speaker separation. \n\nJonathan will also introd
 uce the research topics at the Sagayama/Ono Lab in Tokyo University.\n\nIn
  many problems occuring in musical and acoustical signal processing\, such
  as source separation/localization\, noise cancellation\, multi-pitch anal
 ysis\, harmonic analysis\, rhythm/tempo analysis\, etc.\, there are many c
 ases where ambiguity lies and for which the solution cannot be uniquely de
 termined from the observation. In this presentation\, I will present some 
 work done at the Sagayama/Ono lab on such problems\, mainly with a probabi
 listic approach based on stochastic models.
LOCATION:Engineering Department\, Signal Processing Lab meeting room
END:VEVENT
END:VCALENDAR