Human walking analysis in a random walk framework

Abstract

In this paper, we analyze the human walk movement via a combined deterministic-stochastic model built on top of a random walk framework. By adopting the Laban Effort subcategories to investigate different movement classes, we discuss how stochastic and deterministic forces combine to determine the act of moving and relate the stochastic component to the times at which we can observe a movement change. Major elements of our study are the analysis of random times when there is a walk movement change, the construction of a distribution (waiting time distribution), and the implementation of a dichotomic stochastic force. In particular, we report the observed walk trajectories obtained via an experimental analysis performed in collaboration with a group of volunteers to show, to some degree, the relationship between their walks and the measured relevant random times. Finally, we present our model results via simulated trajectories of a virtual character.