| Title |
Implementation of Application for Smart Healthcare Exercise Management Based on Artificial Intelligence |
| Authors |
하태용(Taeyong Ha) ; 이후진(Hoojin Lee) |
| DOI |
https://doi.org/10.5573/ieie.2020.57.6.44 |
| Keywords |
AI; Deep Learning; PoseNet; Fitness; Healthcare |
| Abstract |
With the spread of the trend of real-time management of personal health using smartphones, and healthcare-related devices are receiving a lot of attention, users are now relying heavily on following the instructor through video content or fitness apps. In this study, the Artificial Intelligence (AI) Convolutional Neural Network (CNN) model, PoseNet, is used to learn pose estimation, analyze human coordinates through this, and implement clapping as part of the motion. We intend to convert the value into the Metabolic Equivalent of Task (MET) and it is also intended to be implemented as a smart healthcare exercise management system application. In particular, PoseNet is effective as a method for determining the calorie consumption of a user's exercise using MET for the motion analysis of real-time pose estimation through a browser, whenever and wherever a device with a camera function is supported. Thus, PoseNet can be considered to be very useful, since it is a model that satisfies the purpose of developing smart healthcare services. |