Computing That Serves

Measuring Human Energy Intake


Thursday, April 9, 2015 - 11:00am


Adam Hoover


Christophe Giraud-Carrier

Colloquium presented by Adam Hoover, Associate Professor at Clemson University
Thursday, April 9, 2015 at 11:00 AM
Location: 1170 TMCB


The technology of wearable sensors has significantly advanced the assessment of energy expenditure in the form of accelerometer-based physical activity monitors.  However, the development of a similar technology for monitoring energy intake has remained elusive.  Previous research has been limited to testing in the laboratory using cumbersome measurements of sound and muscle activity at the throat and ear, and multiple sensors tracking torso and limb movements.  Our group is developing technologies for dietary monitoring using sensors worn on the wrist in the form of watch.  The watch tracks wrist motion all day to automatically detect periods of eating.  We denote each meal or snack detected as an eating activity (EA).  Subsequently, the watch tracks wrist motion during the EA to count the number of bites taken, where a bite is the action of placing food into the mouth.  This talk will present the results of several studies examining the underlying biology of these processes as well as accuracies achieved in detecting EAs and measuring energy intake.  Our work has the potential to replace self-report methods such as food diaries, with the advantages of more objective data, less user burden, and new opportunities for real-time feedback during eating.


Adam Hoover is currently an associate professor in the Electrical and Computer Engineering Department of Clemson University. His research focuses on tracking systems. Tracking can refer to physical problems, such as locating where things are in the world, and also signal problems, such as identifying the relative health of an individual's blood pressure over time. Image and signal processing, mHealth, state space modeling, filtering, and embedded computing form the background of his work. His group works with many types of sensors, and often builds embedded systems that prototype novel tracking ideas.