AI-Assisted yoga Asanas in the future using Deep Learning and Posenet

Yoga instructors have become teaching online due to the ongoing pandemic. Despite studying from top sources including videos, blogs, journals, and essays, users need live tracking to ensure proper posture and health. While technology can aid, beginner-level yoga practitioners rely on their teacher for identifying their proper posture. The main objective of yoga pose detection and corrections is to deliver standard and precise poses for yoga using computer vision. If the yoga pose is not performed correctly, it may lead to severe injuries and long-term problems. Analysing human positions to identify and fix yoga positions can help humans live better lives in their own homes. Our work focuses on experimenting with different approaches to yoga position classification, therefore using PoseNet . Using such algorithms of deep learning, a person can determine the correct/ideal way/method to perform the particular yoga asana he or she is attempting to perform.