Classification of Yoga Pose using Pre-trained CNN Models and Machine Learning Classifiers

In the development of recent technologies over the years have resulted in a widespread growth in the world of Machine learning or Deep learning techniques due to their high performance for several fields or domains of healthcare applications varying from the prediction of health issues such as pneumonia to cricket score predictions, the impressive performance of Deep Learning techniques, have lingering doubts regarding the robustness of Deep Learning in healthcare settings. especially in light of recent results thathave shown that Deep Learning are vulnerable and more prone to attacks such as adversarial attacks. In this paper, we present an overview of various yoga positions and their correct gestures while practicing yoga which could be easily achieved even when one is away from the instructor or yoga tutor for minimizing or prevention of the injuries caused due to incorrect yoga poses. Finally we provide insight into the current research challenges and promising directions for future research using various convolutional network frameworks such as VGG16, VGG19 and Inception V3.