Publication Date: 2022/07/13
Abstract: The most important of yoga poses is known around the world and proves the health benefits preached by ancient sages. As yoga becomes more important, yoga faces the following important challenges: Computer vision technology provides a promising solution for assessing human posture. However, these techniques are rarely used in the areas of health and exercise, and there are no specific references or projects. Named after yoga. This white paper describes the different technologies that can be used for pose estimation and summarize the best ways to use them based on the ease of use of your Android app application. The following describes the methodology used to provide yoga pose estimation in Android applications, how the app is modelled, and how each component works. Pose estimation is a branch of computer vision that deals with the recognition of the individual parts that make up the body (usually the human body). There are several ways to achieve this, The approach I use starts with passing the incoming image through a CNN classifier trained to look for people. When the human body poses are recognized, the pose estimation network searches for trained joints and limbs. The computer can then display the image to the user using markers that identify parts of the body.
Keywords: Deep Learning, Machine Learning, Pose Estimation, OpenPose, PostNet, YogaPoses, CNN.
DOI: https://doi.org/10.5281/zenodo.6824903
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT22MAY619.pdf
REFERENCES