BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Generative Models for Shape and Appearance - Neill Campbell\, Univ
 ersity of Bath
DTSTART:20150626T090000Z
DTEND:20150626T100000Z
UID:TALK59958@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:I will present recent work from SIGGRAPH and CVPR. The first b
 uilds a generative model of fonts in an automatic and unsupervised learnin
 g process\; the only input is a collection of existing font files and the 
 output is a probabilistic manifold that can be used to create new typeface
 s. The second piece of work generalises probabilistic PCA and Active Appea
 rance Models to overcome a fundamental weakness\; existing subspace models
  are unable to model image datasets that cannot be readily aligned. Instea
 d\, we learn a subspace model in a new context space\, a deterministic fun
 ction of an input “part map”\, that implicitly encodes correspondence 
 and thus brings the data into alignment. We illustrate the value of this a
 pproach by considering two example tasks: structured in-painting and appea
 rance transfer.
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR