Application of Artificial Intelligence on Camera-Based Human Pose Prediction for Yoga: A Methodological Study

Due to the physical, mental, and psychological advantages yoga offers the human body, it has evolved into the main activity of the majority of people. Artificial intelligence is developing quickly, and this technology is quickly replacing human labor in a variety of industries, including manufacturing and self-driving cars. This raises the possibility of AI’s development in. A lot of attention has been given to Human Pose Estimation (HPE) in recent years due to the use of deep learning to increase its performance and introduce intriguing new applications, especially in the field of physical activity, bringing us the application of yoga. Reviewing existing research on HPE’s use in yoga is the goal of this study, along with looking at current data and methods, and potential future applications. The paper also presents the performance of the Yoga-82 dataset on the CNN model we have built which achieved the accuracy of 81.59% and compares to previous works. This study and research has been done on the data that is available publicly and has been sourced from online resources and references mentioned.