BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Monte Carlo PHD Filtering - Nick Whiteley\, CUED Signal Processing
  Lab\,
DTSTART:20080612T140000Z
DTEND:20080612T150000Z
UID:TALK11737@talks.cam.ac.uk
CONTACT:Taylan Cemgil
DESCRIPTION:The Probability Hypothesis Density (PHD) filter approximates t
 he optimal filter for a class of dynamical models in which\, at each time\
 , the hidden and observed quantities are spatial point processes. \nSuch m
 odels have applications in multi-object tracking\, audio processing and co
 mmunications engineering\, where the hidden point-process models a time-va
 rying number of unobserved objects\, each of which evolves over time.\n\nO
 riginally formulated in the framework of Finite Set Statistics\, the PHD f
 ilter has the attractive property that it reduces the dimension of the pro
 blem to that of a single unobserved object. However\, in many cases of int
 erest\, the PHD filtering recursion is analytically intractable. This talk
  describes recent advances in the use of Monte Carlo methods to approximat
 e the PHD filter. 
LOCATION:LR12\, Engineering\, Department of
END:VEVENT
END:VCALENDAR