BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Cambridge Image Analysis Seminars
SUMMARY:Below the Surface of the Non-Local Bayesian Image
Denoising Method - Mila Nikolova (speaker) and co-
author Pablo Arias from ENS Cachan
DTSTART;TZID=Europe/London:20170717T130000
DTEND;TZID=Europe/London:20170717T140000
UID:TALK73431AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/73431
DESCRIPTION:The non-local Bayesian (NLB) patch-based approach
of Lebrun\, Buades\, and Morel [1] is considered a
s a state-of-the-art method for the restoration of
(color) images corrupted by white Gaussian noise.
It gave rise to numerous ramifications like e.g.
\, possible improvements\, processing of various d
ata sets and video. This article is the first att
empt to analyse the method in depth in order to un
derstand the main phenomena underlying its effecti
veness. Our analysis\, corroborated by numerical t
ests\, shows several unexpected facts. In a variat
ional setting\, the first-step Bayesian approach
to learn the prior for patches is equivalent to a
pseudo-Tikhonov regularisation where the regularis
ation parameters can be positive or negative. Prac
tically very good results in this step are mainly
due to the aggregation stage - whose importance ne
eds to be re-evaluated.\n\n\nThis is joint work wi
th Pablo Arias.\n\nReference\n[1] Lebrun\, M.\, Bu
ades\, A.\, Morel\, J.M.: A nonlocal Bayesian imag
e denoising algorithm. SIAM J. Imaging Sci.6(3)\,
1665-1688 (2013)
LOCATION:MR 14\, Centre for Mathematical Sciences
CONTACT:Carola-Bibiane Schoenlieb
END:VEVENT
END:VCALENDAR