Review of Artificial Intelligence System For Correcting Exercise Movements and Health Monitoring

Abstract

Abstract: Exercising/yoga/fitness training is an important aspect in everyone's life especially with the modern hectic lifestyle which is very unhealthy. The correctness of an exercise can be judged by certain parameters such as perfect angles between body points and correct breathing techniques and also number of repetitions in some forms of exercise. In yoga, health benefits are achieved only when a person stays in the position for long enough. The Artificial Intelligence system which will help in tracking the movements of user while performing any exercise or yoga asanas will be beneficial. For this system Computer Vision will be used to visualize the body points of the user. Post visualizing the body points, machine learning algorithms are used to detect the posture of the user and check if it is correct or not. Accordingly, the system will generate an output to tell the user the correction to be made in the posture. Also a heartbeat tracking system is used to monitor the heartbeat of the user so that while performing any exercise or yoga asanas if user feels any abnormality in the body it can be tracked by the user's heartbeat and the system could guide the user before any injury occurs. Keywords: CV, CNN, R-CNN, Neural Network, OpenPose, BlazePose, PPG