Scroll to top

Understanding fine-grained details of an individual’s eating pattern using wearable sensors

DECEMBER 4, 2020

Sougata Sen, PhD
Postdoctoral Scholar
Northwestern University

Title: Understanding fine-grained details of an individual’s eating pattern using wearable sensors

About the Presentation: Automatically detecting and monitoring an individual’s eating activity has been a long-standing research goal for both: the mobile sensing research community, as well as for the behavioral health research community. Such automated eating monitoring systems can leverage sensors on wearable devices to support various wellness-related goals – e.g., losing or maintaining target weight, capturing unhealthy eating habits like eating fast, or even detect midnight snacking. Over the years, researchers have evaluated various sensing approaches and applied various machine learning techniques to automatically detect the onset of eating. I will start this talk by describing some of these automated eating activity detection approaches using wearables that we have developed over the years.

Once the eating activity is detected, it is next necessary to monitor the eating activity to understand finer details of such activities — e.g., how many handfuls did the individual consume during a particular meal, or was the individual chewing faster than usual? Such information can enable identification of indicators of problematic eating. I will describe some models that we have developed to detect such fine grained information for an eating activity. Generalized models developed for eating activity monitoring often fail for specific demographic groups. I will also describe how our generalized models fared on specific population subgroups and will describe the advantages and challenges in developing subgroup  specific models.

About the Presenter:  Sougata Sen is a Postdoctoral Scholar at Northwestern University, working with Professor Nabil Alshurafa and Professor Josiah Hester. Prior to joining Northwestern Universty, he was a Postdoctoral Scholar at Dartmouth College, working with Professor David Kotz in topics related to mHealth security and privacy. He was involved in the THaW project. Dr. Sen earned his PhD in 2017 from the School of Information Systems(SIS) at Singapore Management University (SMU), where he was supervised by Professor Archan Misra. Dr. Sen’s PhD thesis was titled “Fusing mobile, wearable and infrastructure sensing for immersive daily lifestyle analytics”. In addition to academic research, Dr. Sen also has more than 6 years of industry research experience, where he was working on topics in the Wireless Sensor Network domain.