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:Isaac Newton Institute Seminar Series
SUMMARY:A Nuclear-norm Model for Multi-Frame Super-resolut
ion Reconstruction - Raymond Chan (Chinese Univers
ity of Hong Kong)
DTSTART;TZID=Europe/London:20171103T111000
DTEND;TZID=Europe/London:20171103T120000
UID:TALK94426AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/94426
DESCRIPTION:In this talk\, we give a new variational approach
to obtain super-resolution images from multiple lo
w-resolution image frames extracted from video cli
ps. First the displacement between the low-resolut
ion frames and the reference frame are computed by
an optical flow algorithm. The displacement matri
x is then decomposed into product of two matrices
corresponding to the integer and fractional displa
cement matrices respectively. The integer displace
ment matrices give rise to a non-convex low-rank p
rior which is then convexified to give the nuclear
-norm regularization term. By adding a standard 2-
norm data fidelity term to it\, we obtain our prop
osed nuclear-norm model. Alternating direction met
hod of multipliers can then be used to solve the m
odel. Comparison of our method with other models o
n synthetic and real video clips shows that our re
sulting images are more accurate with less artifac
ts. It also provides much finer and discernable de
tails. Joint work with Rui Zhao. Research support
ed by HKRGC.
LOCATION:Seminar Room 1\, Newton Institute
CONTACT:INI IT
END:VEVENT
END:VCALENDAR