Virtual sport system for optimum exercising based on a user model

Abstract

It is said that improving movement habits is effective to solve the problem of obesity. In this paper, we describe a physically interactive computer game system that can control a game mode to let a user maintain an appropriate exercise load based on an estimation of the user's heart rate state. We propose a Bayesian network model that can estimate heart rate states. The model calculates probability distribution of a heart rate state using user's motion features. We also describe an experimental validation of the system. © 2009 Springer Berlin Heidelberg.