Decomposition of human motion into dynamics-based primitives with application to drawing tasks

被引：94
作者：
Del Vecchio, D [1 ]
Murray, RM [1 ]
Perona, P [1 ]
机构：
[1] CALTECH, Pasadena, CA 91125 USA
来源：
AUTOMATICA | 2003年 / 39卷 / 12期
关键词：
classification; parameter estimation; learning theory; data acquisition; computer experiments; signal segmentation;
D O I：
10.1016/S0005-1098(03)00250-4
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
Using tools from dynamical systems and systems identification, we develop a framework for the study of primitives for human motion, which we refer to as movemes. The objective is understanding human motion by decomposing it into a sequence of elementary building blocks that belong to a known alphabet of dynamical systems. We develop a segmentation and classification algorithm in order to reduce a complex activity into the sequence of movemes that have generated it. We test our ideas on data sampled from five human subjects who were drawing figures using a computer mouse. Our experiments show that we are able to distinguish between movemes and recognize them even when they take place in activities containing an unspecified number of movemes. (C) 2003 Elsevier Ltd. All rights reserved.
引用
收藏
页码：2085 / 2098
页数：14
相关论文
共 26 条
[1]  
Bishop C. M., 1995, NEURAL NETWORKS PATT
[2]  
BIZZI E, 1999, NEW COGNITIVE NEUROS, P489
[3]   Learning and recognizing human dynamics in video sequences [J].
Bregler, C .
1997 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, PROCEEDINGS, 1997, :568-574
[4]   Introduction to the special section on video surveillance [J].
Collins, RT ;
Lipton, AJ ;
Kanade, T .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2000, 22 (08) :745-746
[5]  
DELVECCHI D, 2002, 2002004 CALTECH
[6]  
DELVECCHIO D, 2003, EUR CONTR C U CAMBR
[7]  
Flanagan JR, 1997, J NEUROSCI, V17, P1519
[8]  
GONCALVES L, 1995, P 7 INT C COMP VIS I, P765
[9]  
GONCALVES L, 1909, 3 INT C AUT FAC GEST, P234
[10]  
Gustafsson F, 2000, ADAPTIVE FILTERING C
123