JOINT BASED FEATURE EXTRACTION FROM MOCAP DATASETS

Accession number;05A0212981
Title;JOINT BASED FEATURE EXTRACTION FROM MOCAP DATASETS
Author; YASUDA H (Tokyo Inst. Technol., Meguro-ku, Jpn) ROMANOS P (Tokyo Inst. Technol., Meguro-ku, Jpn) SAITO S (Tokyo Inst. Technol., Meguro-ku, Jpn) NAKAJIMA M (Tokyo Inst. Technol., Meguro-ku, Jpn)
Journal Title;IEIC Technical Report (Institute of Electronics, Information and Communication Engineers)
Journal Code:S0532B
ISSN:0913-5685
VOL.104;NO.545(IE2004 145-173);PAGE.57-61(2005)
Figure&Table&Reference;FIG.5, TBL.1, REF.10
Pub. Country;Japan
Language;English
Abstract;In the context of video games and characters in the virtual world there has been a great demand for synthesizing realistic motions under explicit user control. In order to meet that demand, a number of methods have been proposed, which typically exploit correlations between adjacent articulation joints. These correlations can in turn be extracted from example motion capture data. In this paper, we examine the relations between the degrees of freedom of each joint individually. We then proceed to show that even when we treat each joint independently, we are able to generate animations that retain the original motion features. (author abst.)