Abstract
Wearable device-based personal activity measurement technology provides various personalized services by integrating bio-signals. However, accurately and rapidly estimating energy expenditure (EE) remains challenging due to user movement and the limitations of measurement parameters. In this paper, we propose Real-Time Energy Expenditure (RTEE), a novel real-time and personalized energy expenditure estimation (EEE) method. The proposed RTEE integrates a Deep Q-Network (DQN)-based activity intensity coefficient inference network with a modified energy consumption prediction algorithm to estimate energy expenditure based on real-time variations in the user's heart rate measurements. Therefore, the proposed algorithm can be applied to various heart rate-based energy consumption prediction methods.