Fitness Help: Yoga Pose Estimation and Correction with Voice Feedback

Yoga, an ancient practice from India, aims to harmonize the body and mind through various physical, mental, and spiritual disciplines. Practitioners often encounter challenges such as physical limitations, injury risks due to improper alignment, and difficulties in maintaining consistency. In response to these challenges, there has been a growing interest in leveraging technology to enhance the yoga experience, with innovations ranging from mobile apps to wearable devices. This paper presents a computational model comprising three stages: realtime data preprocessing, keypoint detection through a MoveNet based CNN model and pose correction through an angle calculation algorithm. The model provides real-time feedback and corrective suggestions, enhancing yoga sessions while minimizing injury risks for a safer and more enjoyable practice by achieving an accuracy rate of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 8 \%}$</tex>.