BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY: Tree-Structured Classifiers for Pose Estimation - Tae-Kyun Kim\, 
 Imperial College
DTSTART:20150720T104500Z
DTEND:20150720T114500Z
UID:TALK60168@talks.cam.ac.uk
CONTACT:Alan Blackwell
DESCRIPTION:Many computer vision tasks can be cast as large-scale classifi
 cation problems\, where extremely efficient and powerful classification me
 thods are pursued. Boosting with decision stump learners\, the state-of-th
 e-art for objet detection\, can be seen as a flat structure\, while many d
 evelopments including a Boosting cascade can be seen as a tree structure. 
 Randomised Decision Forests is an emerging technique in the fields. A hier
 archical structure yields many short paths\, accelerating evaluation time\
 , while feature randomisation promotes good generalisation to unseen data.
  It is inherently for multi-class classification problems. In this talk\, 
 we see applications of Randomised Decision Forests and tree-structured met
 hods with comparisons and insights. The talk focuses on articulated hand p
 ose estimation\, and face recognition/landmarking. Hand and face are highl
 y articulated and deformable objects\, playing a key role for novel man-ma
 chine interfaces. Estimating their 3D postures\, or regressing locations o
 f joints/fiducial points is highly challenging. We have tackled the proble
 ms by various novel ideas on top of the cutting-edge techniques. We conclu
 de the talk with some future directions including active interactive objec
 t recognition.
LOCATION:Lecture Theatre 2\, Computer Laboratory
END:VEVENT
END:VCALENDAR