Title
Human activity modeling as brownian motion on shape manifold
Abstract
In this paper we propose a stochastic modeling of human activity on a shape manifold. From a video sequence, human activity is extracted as a sequence of shape. Such a sequence is considered as one realization of a random process on shape manifold. Then Different activities are modeled by manifold valued random processes with different distributions. To solve the problem of stochastic modeling on a manifold, we first regress a manifold values process to a Euclidean process. The resulted process then could be modeled by linear models such as a stationary incremental process and a piecewise stationary incremental process. The mapping from manifold to Euclidean space is known as a stochastic development. The idea is to parallelly transport the tangent along curve on manifold to a single tangent space. The advantage of such technique is the one to one correspondence between the process in Euclidean space and the one on manifold. The proposed algorithm is tested on database [5] and compared with the related work in [5]. The result demonstrate the high accuracy of our modeling in characterizing different activities.
Year
DOI
Venue
2011
10.1007/978-3-642-24785-9_53
SSVM
Keywords
Field
DocType
stochastic modeling,random process,human activity modeling,shape manifold,manifold values process,human activity,piecewise stationary incremental process,brownian motion,stationary incremental process,euclidean process,different activity,euclidean space
Atlas (topology),Local tangent space alignment,Topology,Center manifold,Closed manifold,Mathematical analysis,Manifold alignment,Invariant manifold,Statistical manifold,Pseudo-Riemannian manifold,Mathematics
Conference
Volume
ISSN
Citations 
6667
0302-9743
2
PageRank 
References 
Authors
0.44
9
3
Name
Order
Citations
PageRank
Sheng Yi120.44
Hamid Krim252059.69
Larry K. Norris371.19