BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20180325T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20171029T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260424T143249Z
UID:59c94b4a10b38833468135@ist.ac.at
DTSTART:20171215T091500
DTEND:20171215T094500
DESCRIPTION:Speaker: Olga Veksler\nhosted by Vladimir Kolmogorov\nAbstract:
  Fully connected pairwise Conditional Random Fields (Full-CRF) with Gaussi
 an edgeweights can achieve superior results compared to sparsely connected
  CRFs. However\, traditional methods for Full-CRFs are too expensive. Prev
 ious work develops efficient approximate optimization based on mean field 
 inference\, which is a local optimization method and can be far from the o
 ptimum. We propose efficient and effective optimization based on graph cut
 s for Full-CRFs with \\em quantized edge weights. To quantize edge weights
 \, we partition the image into superpixels and assume that the weight of a
 n edge between any two pixels depends only on the superpixels these pixels
  belong to. Our quantized edge CRF is an approximation to the Gaussian edg
 e CRF\, and gets closer to it as superpixel size decreases. Being an appro
 ximation\, our model offers an intuition about the regularization properti
 es of the Guassian edge Full-CRF. For efficient inference\, we first consi
 der the two-label case and develop an approximate method based on transfor
 ming the original problem into a smaller domain. Then we handle multi-labe
 l CRF by showing how to implement expansion moves.In both binary and multi
 -label cases\, our solutions have significantly lower energy compared to t
 hat of mean field inference. We also show the effectiveness of our approac
 h on semantic segmentation task.
LOCATION:Mondi Seminar Room 3\, Central Building\, ISTA
ORGANIZER:abonvent@ist.ac.at
SUMMARY:Olga Veksler: Efficient Graph Cut Optimization for Full CRFs with Q
 uantized Edges
URL:https://talks-calendar.ista.ac.at/events/991
END:VEVENT
END:VCALENDAR