BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Mid-level Likelihoods and Constraints for 3D Scene Interpretation 
 - David Fouhey\, The Robotics Institute
DTSTART:20140507T100000Z
DTEND:20140507T110000Z
UID:TALK52241@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:How do you infer the 3D properties of the world from a single 
 image? \nThis question has eluded researchers in psychology and computer v
 ision for decades\, and enabling machines to accomplish this task remains 
 an open question. In this talk\, I will present my research towards solvin
 g the 3D interpretation task.\n\nI will first talk about Data-Driven 3D Pr
 imitives\, a new way of inferring surface normals / scene layout from a si
 ngle image. These primitives are discovered from large-scale RGB-D data by
  optimizing two simple criteria: primitives should be visually discriminat
 ive and geometrically informative. I will show that a straightforward labe
 l-transfer inference approach on top of of these primitives produces state
 -of-the-art results on a complex and cluttered dataset\, as well as effect
 ive cross-dataset performance.\n\nLocal cues\, however\, are inadequate by
  themselves: scenes are highly constrained in structure. I will therefore 
 also talk about my work on constraining the 3D interpretation of scenes vi
 a physical and functional reasoning. Specifically\, I will present work on
  mid-level physical constraints for layout estimation in the form of the c
 onvex and concave edges from the classic line-labeling literature. Finally
 \, I will discuss how recognizing humans in scenes\, even with imprecise a
 nd noisy pose estimators\, can provide valuable cues for scene geometry vi
 a functional reasoning.\n
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR