A Novel Sensor Method for Dietary Detection

A regular diet is closely related to human physical and mental health. Detecting dietary behavior offers a chance to help individuals comprehend their eating habits, promptly recognize any health issues, and adopt suitable measures to enhance their well-being. In this paper, a novel hardware-software sensing method is proposed to enable more people to care about their healthy dietary habits. In this method, the user only needs a speaker-only headphone even without a microphone, typically found in people’s lives to achieve diet detection. Besides, the feasibility of utilizing speaker-only headphones as dietary sensors is investigated through mathematical models and experimental validation and this paper focuses on using this sensor in mastication rate detection. We implement our hardware structure on a 1-layer PCB board and follow the IRB protocol to evaluate its performance, and we achieve MAE (0.865), RMSE (1.26), and average error rate (0.11).