IoT Technologies for HealthCare. 6th EAI International Conference, HealthyIoT 2019, Braga, Portugal, December 4–6, 2019, Proceedings

Research Article

Wi-Fi-Enabled Automatic Eating Moment Monitoring Using Smartphones

Download204 downloads
  • @INPROCEEDINGS{10.1007/978-3-030-42029-1_6,
        author={Zhenzhe Lin and Yucheng Xie and Xiaonan Guo and Chen Wang and Yanzhi Ren and Yingying Chen},
        title={Wi-Fi-Enabled Automatic Eating Moment Monitoring Using Smartphones},
        proceedings={IoT Technologies for HealthCare. 6th EAI International Conference, HealthyIoT 2019, Braga, Portugal, December 4--6, 2019, Proceedings},
        proceedings_a={HEALTHYIOT},
        year={2020},
        month={6},
        keywords={Eating moment monitoring WiFi sensing Healthy eating},
        doi={10.1007/978-3-030-42029-1_6}
    }
    
  • Zhenzhe Lin
    Yucheng Xie
    Xiaonan Guo
    Chen Wang
    Yanzhi Ren
    Yingying Chen
    Year: 2020
    Wi-Fi-Enabled Automatic Eating Moment Monitoring Using Smartphones
    HEALTHYIOT
    Springer
    DOI: 10.1007/978-3-030-42029-1_6
Zhenzhe Lin1,*, Yucheng Xie2,*, Xiaonan Guo2,*, Chen Wang3,*, Yanzhi Ren4,*, Yingying Chen1,*
  • 1: Rutgers University
  • 2: Indiana University-Purdue University Indianapolis
  • 3: Louisiana State University
  • 4: University of Electronic Science and Technology of China
*Contact email: zhenzhe.lin@rutgers.edu, yx11@iupui.edu, xg6@iupui.edu, chenwang1@lsu.edu, renyanzhi05@uestc.edu.cn, yingche@scarletmail.rutgers.edu

Abstract

Dietary habits are closely correlated with people’s health. Study reveals that unhealthy eating habits may cause various diseases such as obesity, diabetes and anemia. To help users create good eating habits, eating moment monitoring plays a significant role. However, traditional methods mainly rely on manual self-report or wearable devices, which either require much user efforts or intrusive dedicated hardware. In this work, we propose a user effort-free eating moment monitoring system by leveraging the WiFi signals extracted from the commercial off-the-shelf (COTS) smartphones. In particular, our system captures the eating activities of users to determine the eating moments. The proposed system can further identify the fine-grained food intake gestures (e.g., eating with fork, knife, spoon, chopsticks and bard hand) to estimate the detailed eating episode for each food intake gesture. Utilizing the dietary information, our system shows the potential to infer the food category and food amount. Extensive experiments with 10 subjects over 400-min eating show that our system can recognize a user’s food intake gestures with up to 97.8% accuracy and estimate the dietary moment within 1.1-s error.