ABSTRACT
Yoga, which originated in ancient times, offers significant physical and mental benefits. However, incorrect postures pose a risk of injury, which highlights the importance of precise guidance during practice. Sage Maharshi Patanjali, who systematized yoga through his scriptures around 500 BC, laid the foundation for modern yoga. In today's digital age, learning yoga has become diverse, from attending classes to using online resources. However, without proper instruction, learners are at risk of performing incorrect postures, leading to injury. This article introduces a deep learning-based system that addresses this challenge by providing real-time feedback and corrections to yoga poses. Combining PoseNet and K-Nearest Neighbours (KNN) algorithms, analyses the user's posture through webcam input, providing personalized guidance to ensure proper posture alignment.
