Embedded system application development on Raspberry Pi 4 – personal A.I trainer (PAT) device for injury prevention

Abstract

The fitness industry has experienced significant growth in recent years, with an increasing number of people signing up for gym memberships. When it comes to working out however, there may be some who struggle with maintaining the proper form, which can lead to injury and reduce the effectiveness of their workouts. To address this issue, this project aims to develop a portable Personal A.I Trainer (PAT) device that uses the MediaPipe pose estimation model to detect key points on the user's body and classify their form. By identifying areas where form correction is needed, the PAT device can help users exercise more safely and effectively.