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:Statistics
SUMMARY:A Framework for Elastic Shape Analysis of Objects
- Anuj Srivastava\, Florida State University
DTSTART;TZID=Europe/London:20150417T160000
DTEND;TZID=Europe/London:20150417T170000
UID:TALK58853AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/58853
DESCRIPTION:Statistical analysis of shapes of objects is an im
portant topic area. Here tools are developed on a
foundation of appropriate mathematical representat
ions and shape metrics.\nWhile a large body of pri
or work is based on registered landmarks\, more co
mplex data demands more general solutions.\nIn thi
s talk we will look at some recent developments\,
with a special focus on the registration problem.\
nRegistration is the process of identifying points
across objects. It not only preserves important\
nstructures in the data but also leads to more par
simonious statistical models for capturing shape\n
variability. In case of parametrized curves and su
rfaces\, the registration step is akin to finding\
noptimal parametrizations. Taking three fundamenta
lly different examples: (1) real-valued function d
ata\,\n(2) parametrized curves in Euclidean spaces
\, and (3) parametrized surfaces in R3\, I will de
scribe\na Riemannian framework that achieves the f
ollowing goals. It provides an analysis of shapes
of objects\nthat is invariant to standard similari
ty transformations and\, additionally\, to paramet
rizations\nof these objects. This framework\, call
ed elastic shape analysis\, incorporates an optima
l registration\nof points across objects while sim
ultaneously providing proper metrics\, geodesics\,
and sample statistics\nof shapes. These sample st
atistics are further useful in statistical modelin
g of shapes in different shape classes. I will dem
onstrate these ideas using applications from medic
al image analysis\, protein structure analysis\, 3
D face recognition\, and human activity recognitio
n in videos.
LOCATION:MR12\, Centre for Mathematical Sciences\, Wilberf
orce Road\, Cambridge
CONTACT:
END:VEVENT
END:VCALENDAR