BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Large-Scale Camera Pose Voting and the Geometric Burstiness Proble
 m - Torsten Sattler\, ETH Zurich
DTSTART:20160419T100000Z
DTEND:20160419T110000Z
UID:TALK65604@talks.cam.ac.uk
CONTACT:44515
DESCRIPTION:Location recognition is the problem of determining the place d
 epicted in a given photo. In the first part of the talk\, we consider the 
 case where the scene is represented by a 3D model and we are not only inte
 rested in the place depicted in an image but also the position and orienta
 tion from which the image was taken\, i.e.\, the camera pose. A major chal
 lenge for solving this image-based localization problem is to establish th
 e 2D-3D matches required for pose estimation. This is especially true for 
 large scale scenes containing many 3D points with similar local appearance
 . Instead of using elaborate matching schemes\, we introduce an efficient 
 camera voting approach whose run-time is independent of the inlier ratio. 
 While our approach allows us to handle an arbitrary number of matches in l
 inear time\, we show that simply increasing the number of 2D-3D matches us
 ed for pose estimation does not solve the image-based localization problem
 . The reason for this behavior is that increasing the number of matches le
 ads to wrong poses with many inliers. In the second part of the talk\, we 
 thus consider the problem of finding a better decision criterion than the 
 raw inlier count. We show that geometric bursts\, i.e.\, spatial configura
 tions appearing at multiple places in a scene\, are a major reason why the
  raw inlier count fails for large-scale location recognition. We introduce
  simple schemes that allow us to efficiently detect geometric bursts durin
 g query time. We show experimentally that down-weighting inliers based on 
 the number of bursts they appear in allows us to better decide between cor
 rect and incorrect place recognition results and significantly boosts the 
 location recognition performance.\n\n
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR