BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Dimension reduction into superpixels for image analysis and proces
 sing - Rémi Giraud 
DTSTART:20190311T110000Z
DTEND:20190311T120000Z
UID:TALK121588@talks.cam.ac.uk
CONTACT:AI Aviles-Rivero
DESCRIPTION:Superpixels locally group pixels into regions and have become 
 very popular in image processing and computer vision applications. The aim
  is to exploit the local redundancy of information to lower the computatio
 nal burden and to potentially improve the performances by reducing the noi
 se of a processing at the pixel level. Superpixels can also be considered 
 as a multi-resolution app roach that preserves image contours\, contrary t
 o standard regular down-sampling methods.\nThe presentation will first foc
 us on the superpixel decomposition process itself. We answer to the limita
 tions of state-of-the-art methods that may still highly fail to produce ac
 curate pixel clustering in the presence of object contours\, noise\, or te
 xtures.\nThen\, we propose a new framework adapting standard regular patch
 -based approaches\, e.g.\, non-local means\, to these irregular regions. R
 esults on several applications such as segmentation\, classification and c
 olor transfer will be presented.
LOCATION:MR13 Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR