BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:An Approximate Differentiable Renderer - Matthew Loper\; Max Planc
 k Institute for Intelligent Systems
DTSTART:20141020T090000Z
DTEND:20141020T100000Z
UID:TALK55638@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Although computer vision can be posed as an inverse rendering 
 problem\, most renderers are not tailored to this task. Our framework make
 s it simple to express a forward graphics model\, automatically obtain der
 ivatives with respect to the model parameters\, and optimize over them. Bu
 ilt on a new autodifferentiation package and OpenGL\, OpenDR provides colo
 r\, depth and boundary renderers\, and promotes the ability to construct f
 unctions of the renderers. We demonstrate the power and simplicity of prog
 ramming with OpenDR by using it to prototype a system for estimating human
  body shape from Kinect depth and RGB data.
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR