Abstract: This paper presents system based on pose detection for yoga pose recognition. Yoga, which has its roots in India, is a form of physical activity that helps your body and mind work together properly for each of their individual functions. In many nations around the world, yoga of Indian heritage is primarily used to maintain health. As a result, the body's posture while practising yoga has a significant impact on one's health. Many medical professionals advised doing yoga to help patients recover from injuries more quickly and as the greatest way to combat mental health issues like depression, anxiety, and post-traumatic stress. Some yoga practitioners don't do their yoga postures correctly, which causes a variety of physical issues such joint discomfort, disc difficulties, shoulder pain, etc. We are launching a tool for categorising and improving your yoga. It will use MoveNet as a classification model to examine your yoga-related bodily motions.
Keywords: Non-Parametric Weight Feature Extraction (NFWE), Principal Component Analysis (PCA), TensorFlow, MoveNet, 17-Keypoints.
|
DOI:
10.17148/IJARCCE.2023.12491
