Human Exercise Detection and Correction using Media Pipe Framework

Fitness activities offer numerous health benefits, but improper execution can lead to inefficiency and potential harm. When performing exercises incorrectly, individuals often fail to maintain the correct form or posture. In this research, a program has been developed to assess and provide feedback on the user's workout posture. Utilizing a camera, this system offers realtime interaction by detecting joints and evaluating workout accuracy through vector angles. The implementation leverages MediaPipe, a cross- platform ML framework compatible with Windows and Linux computers equipped with a webcam.