Nonparametric Learning in Computer Graphics
開催期間
13:00 ~ 15:00
場所
講演者
概要
Cheap computer storage and processing has enabled practical "nonparametric"
statistical learning methods. This talk will briefly survey nonparametric
learning and present applications showing how labor-intensive artistic
problems in computer graphics can be accelerated with these
techniques. Specific topics will include nonparametric modeling of character
motion, priors for rotoscoping and facial animation, and modeling of facial
appearance.
J.P. Lewis is a research lead at Weta Digital and a Senior Lecturer at
Victoria University of Wellington. Lewis has software credits on several
films including Avatar, The Matrix sequels, and Forrest Gump. His published
algorithms have been incorporated in commercial software and are widely
adopted in the film industry. His current work involves applications of
computer vision and machine learning techniques to special effects.