BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Fast noise learning via nonlinear PDE constrained optimization - L
 uca Calatroni (CCA\, CIA)
DTSTART:20131127T160000Z
DTEND:20131127T170000Z
UID:TALK46772@talks.cam.ac.uk
CONTACT:Marcus Webb
DESCRIPTION:In this talk\, we recap the framework of Bounded Variation fun
 ctions and their main features relevant to Imaging.  In the course of that
  discussion we highlight some drawbacks of TV reconstruction and introduce
  higher-order versions of TV which improve upon them\, highlighting the nu
 merical obstacles arising when solving such models and possible directions
  (such as ADI directional splitting schemes).\n\nThe rest of talk will foc
 us on my current research on the optimal setup of these models by means a 
 nonlinear PDE constrained optimisation\, based on a training set of origin
 al and noisy images. In such approach\, the optimal weights are computed s
 uch that the corresponding total variation regularised solutions (encoded 
 in the constraints) 'best' fit the original images. To improve upon the ef
 ficiency of the numerical solvers\, we use dynamical sampling schemes to c
 ompute the optimal parameters.\n\nThis is a joint work with J. C. De Los R
 eyes and C.-B. Schönlieb.
LOCATION:MR14\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR