Recognizing Drinking Gestures with Wrist-Worn Inertial Sensors: Public Dataset and Lightweight CNN Approach

• Pedro D. Gohl UFAM
• Laura I. Queiroz UFAM
• Eduardo Souto UFAM
• Eulanda M. Santos UFAM

DOI: https://doi.org/10.5753/sbcas.2026.21744

Resumo

Monitoring fluid intake via wearable sensors is challenging due to gesture variability and overlap with similar hand movements. We propose a deep learning approach for liquid intake detection using wrist-worn accelerometer and gyroscope data, segmented into 4-second windows at 50 Hz. We introduce a Multi-Scale 1D Convolutional Neural Network (MS-Conv1D) that extracts spatiotemporal features at multiple temporal resolutions. Evaluated via 50-fold Leave-One-Subject-Out cross-validation, the model achieved F1-scores of 93.33% (binary) and 91.50% (multiclass), outperforming traditional classifiers and most state-of-the-art baselines, while remaining efficient enough for real-time execution on embedded hardware.

Referências

Amft, O., Bannach, D., Pirkl, G., Kreil, M., and Lukowicz, P. (2010). Towards wearable sensing-based assessment of fluid intake. In 2010 8th IEEE International Conference on Pervasive Computing and Communications Workshops (PERCOM Workshops), pages 298–303.

Anderez, D. O., Lotfi, A., and Langensiepen, C. (2018). A hierarchical approach in food and drink intake recognition using wearable inertial sensors. In Proceedings of the 11th Pervasive Technologies Related to Assistive Environments Conference, pages 552–557.

Begg, D. P. (2017). Disturbances of thirst and fluid balance associated with aging. Physiology & Behavior, 178:28–34.

Cergolj, V. et al. (2024). Drinking event detection on a sensing wristband using machine learning. Journal of Ambient Intelligence and Smart Environments, pages 1–20. Preprint.

Cohen, R., Fernie, G., and Roshan Fekr, A. (2021). Fluid intake monitoring systems for the elderly: a review of the literature. Nutrients, 13(6):2092.

Gomes, D. and Sousa, I. (2019). Real-time drink trigger detection in free-living conditions using inertial sensors. Sensors.

Griffith, H. and Biswas, S. (2019). Improving water consumption estimates from a bottle-attachable sensor using heuristic fusion. In 2019 IEEE 20th International Symposium on A World of Wireless, Mobile and Multimedia Networks (WoWMoM), pages 1–3.

Hamatani, T., Elhamshary, M., Uchiyama, A., and Higashino, T. (2018). Fluidmeter: Gauging the human daily fluid intake using smartwatches. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, pages 1–25.

Huang, H.-Y. et al. (2020). Fluid intake monitoring system using a wearable inertial sensor for fluid intake management. Sensors, 20(22):6682.

Iosifidis, A., Marami, E., Tefas, A., and Pitas, I. (2012). Eating and drinking activity recognition based on discriminant analysis of fuzzy distances and activity volumes. In 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 2201–2204.

Li, J.-H. et al. (2024). Multi-sensor fusion approach to drinking activity identification for improving fluid intake monitoring. Applied Sciences, 14(11):4480.

Liaqat, S., Dashtipour, K., Rizwan, A., Usman, M., Shah, S. A., Arshad, K., Assaleh, K., and Ramzan, N. (2022). Personalized wearable electrodermal sensing-based human skin hydration level detection for sports, health and wellbeing. Scientific Reports.

Likhtenshtein, G. I. (2021). Water: Clinical aspects. In Biological Water: Physicochemical Aspects, pages 481–512. unknown.

Liu, K.-C., Hsieh, C.-Y., Huang, H.-Y., Chiu, L.-T., Hsu, S. J.-P., and Chan, C.-T. (2020). Drinking event detection and episode identification using 3d-printed smart cup. IEEE Sensors Journal, pages 13743–13751.

Lorenzo, I., Serra-Prat, M., and Yébenes, J. C. (2019). The role of water homeostasis in muscle function and frailty: A review. Nutrients.

Martínez, P., Gordillo-Castillo, N., and Cortés Sáenz, D. (2023). Towards fluid intake quantification in older adults: An algorithm for movement detection using accelerometry and gyroscope sensors. In Congreso Nacional de Ingeniería Biomédica, pages 222–231, Cham. Springer Nature Switzerland.

Moccia, S., Solbiati, S., Khornegah, M., Bossi, F. F., and Caiani, E. G. (2022). Automated classification of hand gestures using a wristband and machine learning for possible application in pill intake monitoring. Computer Methods and Programs in Biomedicine, 219:106753.

Roy, A., Dutta, H., Griffith, H., and Biswas, S. (2022). An on-device learning system for estimating liquid consumption from consumer-grade water bottles and its evaluation. Sensors, 22(7):2514.

Senyurek, V., Imtiaz, M., and Hassan, N. (2019). Detection of drinking via a wrist-worn inertial sensor.

Tham, J. S., Chang, Y. C., and Fauzi, M. F. A. (2014). Automatic identification of drinking activities at home using depth data from rgb-d camera. In The 2014 International Conference on Control, Automation and Information Sciences (ICCAIS 2014), pages 153–158. IEEE.

Wang, C., Kumar, T. S., De Raedt, W., Camps, G., Hallez, H., and Vanrumste, B. (2022). Drinking gesture detection using wrist-worn imu sensors with multi-stage temporal convolutional network in free-living environments. In 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), pages 1778–1782.

Wang, C., Kumar, T. S., De Raedt, W., Camps, G., Hallez, H., and Vanrumste, B. (2024). Eating speed measurement using wrist-worn imu sensors towards free-living environments. IEEE journal of biomedical and health informatics, 28(10):5816–5828.

WitMotion (2024). WT901BLECL 9-axis BLE accelerometer inclinometer. [link].