Detecting different physical exercises’ postures using deep learning with MediaPipe and OpenCV

This abstract has open access
Abstract Summary/Description
The primary aim of this paper is to delve into the enhancement of body posture during exercise utilizing an innovative AI-based smart system. This system is designed to provide recommendations for improved posture by employing real-time images and video sensing technology. Acknowledging the profound impact of mental health on our daily routines and physical workout regimens, this paper underscores the critical importance of regular physical exercise in maintaining optimal hormone levels and fostering mindfulness. It underscores the the necessity of executing physical practices safely and properly to prevent potential harm to the body. Consequently, continual monitoring and adjustment of exercise techniques are advocated to ensure their effectiveness and safety. The study focuses on exploring AI-based exercise monitoring systems, leveraging various Python modules such as MediaPipe, TensorFlow, Matplotlib, and OpenCV, among others. These modules are instrumental in analyzing input data comprehensively and furnishing precise feedback on body posture, thereby facilitating informed adjustments to exercise routines. Keywords:- Physical exercise, OpenCV, TensorFlow, CNN, Media Pipe.
Abstract ID :
NKDR205
Georgia State University, College Of Arts And Sciences
2 visits