An On-Device Learning System for Estimating Liquid Consumption from Consumer-Grade Water Bottles and Its Evaluation
dc.contributor.author | Roy, Avirup | |
dc.contributor.author | Dutta, Hrishikesh | |
dc.contributor.author | Griffith, Henry | |
dc.contributor.author | Biswas, Subir | |
dc.date.accessioned | 2022-04-11T13:59:24Z | |
dc.date.available | 2022-04-11T13:59:24Z | |
dc.date.issued | 2022-03-25 | |
dc.date.updated | 2022-04-11T13:59:26Z | |
dc.description.abstract | A lightweight on-device liquid consumption estimation system involving an energy-aware machine learning algorithm is developed in this work. This system consists of two separate on-device neural network models that carry out liquid consumption estimation with the result of two tasks: the detection of sip from gestures with which the bottle is handled by its user and the detection of first sips after a bottle refill. This predictive volume estimation framework incorporates a self-correction mechanism that can minimize the error after each bottle fill-up cycle, which makes the system robust to errors from the sip classification module. In this paper, a detailed characterization of sip detection is performed to understand the accuracy-complexity tradeoffs by developing and implementing a variety of different ML models with varying complexities. The maximum energy consumed by the entire framework is around 119 mJ during a maximum computation time of 300 µs. The energy consumption and computation times of the proposed framework is suitable for implementation in low-power embedded hardware that can be incorporated in consumer grade water bottles. | |
dc.description.department | Electrical and Computer Engineering | |
dc.identifier | doi: 10.3390/s22072514 | |
dc.identifier.citation | Sensors 22 (7): 2514 (2022) | |
dc.identifier.uri | https://hdl.handle.net/20.500.12588/819 | |
dc.rights | Attribution 4.0 United States | |
dc.rights.uri | https://creativecommons.org/licenses/by/4.0/ | |
dc.subject | drink volume estimation | |
dc.subject | sip detection | |
dc.subject | embedded machine learning | |
dc.subject | on-device classification | |
dc.subject | neural networks | |
dc.subject | TinyML | |
dc.title | An On-Device Learning System for Estimating Liquid Consumption from Consumer-Grade Water Bottles and Its Evaluation | |
dc.type | Article |