Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Joint Modeling of Human Energy Expenditure and Physical Activity Using Wearable Sensors
How can knowledge regarding physical activity type impact the estimation of human energy expenditure using wearable sensors?
Human energy expenditure (EE) refers to the amount of energy an individual uses to maintain essential body functions (respiration, circulation, digestion) and because of physical activity. Knowledge regarding the expended energy or calories could help people (e.g., athletes, obese, diabetic) to plan their physical activity for leading a healthier lifestyle. Additionally, it could be used to enable nutrition coaching for weight management purposes. Devising accurate methods for EE estimation is a key enabler of the mentioned intervention strategies and it is the core goal of this project.
In this project, we will use an existing dataset, which has been collected while participants performed different physical activities (e.g., cycling, running). The dataset contains sensor data collected using seven wearable devices placed on four body locations (head, ear, chest, and wrist) and respiratory data collected with an indirect calorimeter as well as other information (e.g., demographics and body composition data, type and intensity of physical activity performed). The goal of the project is to develop an approach to use sensor data to estimate human energy expenditure and recognize the physical activity. The student will be asked to implement single-and multi-task pipelines to investigate the performance of separate and joint learning of human energy expenditure and physical activity.
Keywords: multitask learning, wearable sensing, energy expenditure estimation, physical activity recognition, signal processing
Not specified
Not specified
Shkurta Gashi, shkurta.gashi@ai.ethz.ch,
Prof. Dr. Christian Holz, christian.holz@inf.ethz.ch
Shkurta Gashi, shkurta.gashi@ai.ethz.ch, Prof. Dr. Christian Holz, christian.holz@inf.ethz.ch