HydraDoctor: real-time liquids intake monitoring by collaborative sensing

Water has been widely acknowledged as an essential part of all living things. It is the fundamental necessity for all life's activities and most biochemical reactions in human body are executed in water. Therefore, the type and quantity of liquid intake everyday have a critical impact on individuals' health. In this paper, we demonstrate HydraDoctor, a real-time liquids intake monitoring system which is able to detect drinking activities, classify the categories of liquids and estimate the amount of intake. The system runs on multiple platforms including a smartwatch to detect the motion of hands and a smartglass to capture the images of mugs. A smartphone is also used as an edge computing platform and a remote server is designed for computationally intensive image processing. In HydraDoctor, multiple state-of-the-art machine learning techniques are applied: a Support Vector Machine (SVM)-based classifier is proposed to achieve accurate and efficient liquids intake monitoring, which is trained to detect the hand raising action. Both of them are well optimized to enable in-situ processing on smartwatch. To provide more robust and detailed monitoring, the smartglass is also incorporated and trigged to capture a short video clip in the front of the user when potential drinking activity is detected. The smartglass will send the video clip to the remote server via its companion smartphone and a Faster-RCNN is performed on the server to confirm the detected drinking activity and identify the type of intake liquid. According to our evaluation on the real-world experiments, HydraDoctor achieves very high accuracy both in drinking activity detection and types of liquids classification, whose accuracy is 85.64% and 84% respectively.

[1]  Wei Wang,et al.  Understanding and Modeling of WiFi Signal Based Human Activity Recognition , 2015, MobiCom.

[2]  Wang Yi,et al.  AutoDietary: A Wearable Acoustic Sensor System for Food Intake Recognition in Daily Life , 2016, IEEE Sensors Journal.

[3]  Damith Chinthana Ranasinghe,et al.  Real-time fluid intake gesture recognition based on batteryless UHF RFID technology , 2017, Pervasive Mob. Comput..

[4]  Gregory D. Abowd,et al.  A practical approach for recognizing eating moments with wrist-mounted inertial sensing , 2015, UbiComp.

[5]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Paul Lukowicz,et al.  Towards wearable sensing-based assessment of fluid intake , 2010, 2010 8th IEEE International Conference on Pervasive Computing and Communications Workshops (PERCOM Workshops).

[7]  Parisa Rashidi,et al.  Power-efficient real-time approach to non-wear time detection for smartwatches , 2017, 2017 IEEE EMBS International Conference on Biomedical & Health Informatics (BHI).

[8]  Ju Wang,et al.  TagScan: Simultaneous Target Imaging and Material Identification with Commodity RFID Devices , 2017, MobiCom.

[9]  Circadian variation and responsiveness of hydration biomarkers to changes in daily water intake , 2013, European Journal of Applied Physiology.

[10]  Majid Sarrafzadeh,et al.  A Wearable Nutrition Monitoring System , 2014, 2014 11th International Conference on Wearable and Implantable Body Sensor Networks.

[11]  Jake K. Aggarwal,et al.  Human activity recognition from 3D data: A review , 2014, Pattern Recognit. Lett..

[12]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[13]  Sen Wang,et al.  Snoopy , 2018, Proc. ACM Interact. Mob. Wearable Ubiquitous Technol..

[14]  Fei-Fei Li,et al.  ImageNet: A large-scale hierarchical image database , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[15]  S. Shirmohammadi,et al.  A novel method for measuring nutrition intake based on food image , 2012, 2012 IEEE International Instrumentation and Measurement Technology Conference Proceedings.

[16]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Paul Lukowicz,et al.  Smart-surface: Large scale textile pressure sensors arrays for activity recognition , 2016, Pervasive Mob. Comput..

[18]  Wolf-Joachim Fischer,et al.  Food intake monitoring: an acoustical approach to automated food intake activity detection and classification of consumed food , 2012, Physiological measurement.

[19]  Zhen Wang,et al.  Draining our glass: an energy and heat characterization of Google Glass , 2014, APSys.

[20]  S M Kleiner,et al.  Water: an essential but overlooked nutrient. , 1999, Journal of the American Dietetic Association.

[21]  Weihua Sheng,et al.  AutoHydrate: A wearable hydration monitoring system , 2016, 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).