BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:5506@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20130606T100000
DTEND;TZID=Europe/Paris:20130606T110000
DTSTAMP:20241029T171220Z
URL:https://www.i2m.univ-amu.fr/evenements/p-vandergheynst-epfl-robust-ima
 ge-reconstruction-from-multi-view-measurements/
SUMMARY: (...): P. Vandergheynst (EPFL): Robust image reconstruction from m
 ulti-view measurements
DESCRIPTION:: Robust image reconstruction from multi-view measurements.\nBy
  Pierre Vandergheynst\\\, EPFL.\n\nWe propose a novel method to accurately
  reconstruct a set of images representing a single scene from few linear m
 ulti-view measurements. Each observed image is modeled as the sum of a bac
 kground image and a foreground one. The background image is common to all 
 observed images but undergoes geometric transformations\\\, as the scene i
 s observed from dierent viewpoints. In this paper\\\, we assume that these
  geometric transformations\nare represented by a few parameters\\\, e.g.\\
 \, translations\\\, rotations\\\, ane transformations\\\, etc.. The foregr
 ound images dier from one observed image to another\\\, and are used to mo
 del possible occlusions of the scene. The proposed reconstruction algorith
 m estimates jointly the images and the transformation parameters from the 
 available multi-view measurements. The ideal solution of this multi-view i
 maging problem minimizes a non-convex functional\\\, and the reconstructio
 n technique is an alternating descent method built to minimize this functi
 onal. The convergence of the proposed algorithm is studied\\\, and conditi
 ons under which the sequence of estimated images and parameters converges 
 to a critical point of the non-convex functional are provided. Finally\\\,
  the eciency of the algorithm is demonstrated using numerical simulations 
 for applications such as compressed sensing or super-resolution.
CATEGORIES:Séminaire,Signal et Apprentissage
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20130331T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR