SmartGym body weight sensor

Abstract

Regular partaking of Leisure-Time Physical Activity (LTPA) is essential for a healthy life. LTPA includes exercises and sports. With the added benefits of instant gratifications, gamification can make exercises more enjoyable. Interactive gameplay incorporated with workouts can motivate and engage individuals to exercise more often. Targeted to gamify callisthenics workouts, sit-up serves as the first bodyweight workout for the proof of concept. Exercise detection and repetition counting were applied as mechanisms for the game control. This was made possible through key-points detection using computer vision and MoveNet. This project sought to support the endeavour of gamifying bodyweight workouts by setting up the infrastructure needed. In the past year, a replicate of the Minimum Viable Product (MVP) of the bodyweight station was set up. This allowed for development and testing to be done for the new features. New features included the Mode Selector, Key-point Service (KPS), and an additional camera pipeline. Integration of work was done with the project titled SmartGym Bodyweight Gamification which built the game portion of this endeavour. Furthermore, this project attempted to collect a workout dataset to train an additional workout for detection. However, this was cut short due to time constraints. Despite challenges faced due to global chip shortage, deployment was still done on a desktop as the next best alternative. Overall, this project achieved the aim of setting up the necessary services for gamification.