Measuring motion significance and motion complexity

详细信息    查看全文

• 作者：Il Hong Suh^a ; ^{ihsuh@hanyang.ac.kr} ; Sang Hyoung Lee ; ^b ; ^{zelog@kitech.re.kr} ; Nam Jun Cho^a ; ^{namjun@hanyang.ac.kr} ; Woo Young Kwon^c ; ^{wykwon@wisenut.co.kr}
• 关键词：Motion significance ; Motion complexity ; Temporal entropy ; Spatial entropy ; Gaussian mixture model
• 刊名：Information Sciences
• 出版年：2017
• 出版时间：May 2017
• 年：2017
• 卷：388-389
• 期：Complete
• 页码：84-98
• 全文大小：2690 K
• 卷排序：388

文摘

In this paper, we propose two novel measures to specify motion significance and motion complexity from human motion trajectories. Motion significance indicates the relative meaningfulness of every motion frame which is defined as a set of data points acquired at a time index from multiple motion trajectories. Motion complexity indicates the number of meaningful motion frames involved in a set of such human motions. For this, we first show that motion significance can be measured by considering both temporal entropy and spatial entropy of a motion frame, based on the analysis of Gaussian mixtures learned from human motions. Motion complexity is then calculated by measuring the averaged amount of motion significance involved in all time indexes of motion trajectories. These two measures are devised to satisfy the requirement of neural complexity measure proposed to attain small values for totally random or totally regular activities. To show that the proposed measures are consistent with our intuitive notion of motion significance and motion complexity, several human motions for drawing and pouring are analyzed by means of motion significance and motion complexity. Furthermore, our complexity measure is compared with three existing complexity measures to analyze their similarity and dissimilarity.