BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Epitomic representation of natural signals - Anitha Kannan\, Micro
 soft Research Cambridge
DTSTART:20060322T150000Z
DTEND:20060322T160000Z
UID:TALK4824@talks.cam.ac.uk
CONTACT:Phil Cowans
DESCRIPTION:Many natural signals such as images\, audio and genetic sequen
 ces have repeating fragments that can be compactly expressed using\, what 
 we call as\, an epitome. An epitome of a signal is its compact representat
 ion that is learned using overlapping segments or patches of various sizes
  from the signal.  In this talk\, I will explain a generative model for im
 ages using epitomes. When the epitomic representation is used in a hierarc
 hical generative model of image formation\, appropriate inference algorith
 ms can be derived to extract epitomes from a single image or a collection 
 of images and simultaneously perform various tasks such as image segmentat
 ion\, motion estimation\, super-resolution and image denoising. \nTime per
 mitting\, I will preview recent work by colleagues on extending this to vi
 deos.\n\nJoint work with  Nebojsa Jojic (Microsoft Research) and Brendan F
 rey (University of Toronto) 
LOCATION:Ryle Seminar Room\, Cavendish Laboratory
END:VEVENT
END:VCALENDAR