Energy expenditure estimation DEMO application

Abstract

The paper presents two prototypes for the estimation of human energy expenditure during normal daily activities and exercise. The first prototype employs two dedicated inertial sensors attached to the user's chest and thigh and a heart rate monitor. The second prototype uses only the accelerometer embedded in a smart phone carried in the user's pocket. Both systems use machine learning for the energy expenditure estimation. The focus of the demo is the convenience of using a smart phone application to provide the user with real-time insight into his/hers current status of the expended energy and also for on-the-spot encouragement based on the status. The evaluation and validation of both systems were done against the Cosmed indirect calorimeter, a gold standard for energy expenditure estimation and against the SenseWear, a dedicated commercial product for energy expenditure estimation. © 2013 Springer International Publishing Switzerland.