A user-adaptive modeling for eating action identification from wristband time series

Junghyo Lee, Prajwal Paudyal, Ayan Banerjee, Sandeep K.S. Gupta

Research output: Contribution to journalArticle

Abstract

Eating activity monitoring using wearable sensors can potentially enable interventions based on eating speed to mitigate the risks of critical healthcare problems such as obesity or diabetes. Eating actions are polycomponential gestures composed of sequential arrangements of three distinct components interspersed with gestures that may be unrelated to eating. This makes it extremely challenging to accurately identify eating actions. The primary reasons for the lack of acceptance of state-of-the-art eating action monitoring techniques include the following: (i) the need to install wearable sensors that are cumbersome to wear or limit the mobility of the user, (ii) the need for manual input from the user, and (iii) poor accuracy in the absence of manual inputs. In this work, we propose a novel methodology, IDEA, that performs accurate eating action identification within eating episodes with an average F1 score of 0.92. This is an improvement of 0.11 for precision and 0.15 for recall for the worst-case users as compared to the state of the art. IDEA uses only a single wristband and provides feedback on eating speed every 2 min without obtaining any manual input from the user.

Original languageEnglish (US)
Article number22
JournalACM Transactions on Interactive Intelligent Systems
Volume9
Issue number4
DOIs
StatePublished - Oct 2019

Fingerprint

Time series
Monitoring
Medical problems
Wear of materials
Feedback
Wearable sensors

Keywords

  • Diet monitoring
  • Gesture recognition
  • Time-series data modeling
  • Wearable

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Artificial Intelligence

Cite this

A user-adaptive modeling for eating action identification from wristband time series. / Lee, Junghyo; Paudyal, Prajwal; Banerjee, Ayan; Gupta, Sandeep K.S.

In: ACM Transactions on Interactive Intelligent Systems, Vol. 9, No. 4, 22, 10.2019.

Research output: Contribution to journalArticle

@article{c37afd5ebcd2458da921dc22284578cb,
title = "A user-adaptive modeling for eating action identification from wristband time series",
abstract = "Eating activity monitoring using wearable sensors can potentially enable interventions based on eating speed to mitigate the risks of critical healthcare problems such as obesity or diabetes. Eating actions are polycomponential gestures composed of sequential arrangements of three distinct components interspersed with gestures that may be unrelated to eating. This makes it extremely challenging to accurately identify eating actions. The primary reasons for the lack of acceptance of state-of-the-art eating action monitoring techniques include the following: (i) the need to install wearable sensors that are cumbersome to wear or limit the mobility of the user, (ii) the need for manual input from the user, and (iii) poor accuracy in the absence of manual inputs. In this work, we propose a novel methodology, IDEA, that performs accurate eating action identification within eating episodes with an average F1 score of 0.92. This is an improvement of 0.11 for precision and 0.15 for recall for the worst-case users as compared to the state of the art. IDEA uses only a single wristband and provides feedback on eating speed every 2 min without obtaining any manual input from the user.",
keywords = "Diet monitoring, Gesture recognition, Time-series data modeling, Wearable",
author = "Junghyo Lee and Prajwal Paudyal and Ayan Banerjee and Gupta, {Sandeep K.S.}",
year = "2019",
month = "10",
doi = "10.1145/3300149",
language = "English (US)",
volume = "9",
journal = "ACM Transactions on Interactive Intelligent Systems",
issn = "2160-6455",
publisher = "Association for Computing Machinery (ACM)",
number = "4",

}

TY - JOUR

T1 - A user-adaptive modeling for eating action identification from wristband time series

AU - Lee, Junghyo

AU - Paudyal, Prajwal

AU - Banerjee, Ayan

AU - Gupta, Sandeep K.S.

PY - 2019/10

Y1 - 2019/10

N2 - Eating activity monitoring using wearable sensors can potentially enable interventions based on eating speed to mitigate the risks of critical healthcare problems such as obesity or diabetes. Eating actions are polycomponential gestures composed of sequential arrangements of three distinct components interspersed with gestures that may be unrelated to eating. This makes it extremely challenging to accurately identify eating actions. The primary reasons for the lack of acceptance of state-of-the-art eating action monitoring techniques include the following: (i) the need to install wearable sensors that are cumbersome to wear or limit the mobility of the user, (ii) the need for manual input from the user, and (iii) poor accuracy in the absence of manual inputs. In this work, we propose a novel methodology, IDEA, that performs accurate eating action identification within eating episodes with an average F1 score of 0.92. This is an improvement of 0.11 for precision and 0.15 for recall for the worst-case users as compared to the state of the art. IDEA uses only a single wristband and provides feedback on eating speed every 2 min without obtaining any manual input from the user.

AB - Eating activity monitoring using wearable sensors can potentially enable interventions based on eating speed to mitigate the risks of critical healthcare problems such as obesity or diabetes. Eating actions are polycomponential gestures composed of sequential arrangements of three distinct components interspersed with gestures that may be unrelated to eating. This makes it extremely challenging to accurately identify eating actions. The primary reasons for the lack of acceptance of state-of-the-art eating action monitoring techniques include the following: (i) the need to install wearable sensors that are cumbersome to wear or limit the mobility of the user, (ii) the need for manual input from the user, and (iii) poor accuracy in the absence of manual inputs. In this work, we propose a novel methodology, IDEA, that performs accurate eating action identification within eating episodes with an average F1 score of 0.92. This is an improvement of 0.11 for precision and 0.15 for recall for the worst-case users as compared to the state of the art. IDEA uses only a single wristband and provides feedback on eating speed every 2 min without obtaining any manual input from the user.

KW - Diet monitoring

KW - Gesture recognition

KW - Time-series data modeling

KW - Wearable

UR - http://www.scopus.com/inward/record.url?scp=85075631112&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075631112&partnerID=8YFLogxK

U2 - 10.1145/3300149

DO - 10.1145/3300149

M3 - Article

AN - SCOPUS:85075631112

VL - 9

JO - ACM Transactions on Interactive Intelligent Systems

JF - ACM Transactions on Interactive Intelligent Systems

SN - 2160-6455

IS - 4

M1 - 22

ER -