Organizing motions with a nominal description

Due to the success of motion capture technologies, large motion capture data becomes available. Although organizing large databases has been widely researched for various purposes, there is little attention to motion capture data. In this paper, we propose an effective scheme to organize large motion databases according to its nominal features. To do this, we first define a set of attributes and its corresponding values to form a set of attribute-value pairs. Each attribute represents a visual characteristic of human motion. We adopt machine learning algorithms to assign a specific value to the attribute. Then, we categorize all motions in database according to the set of attribute-value pairs. Since attributes define nominal properties rather than numerical 1 

• Type: Journal
• Year: 2007
• Where: JVCA
• Authors: Min Je Park, Jieun Cho
• Keywords: JVCA 2007, Large Motion, Motion Capture, Motion Capture Data