A REVIEW ON: YOGA POSE DETECTION USING DEEP LEARNING
Prof. Pravin M. Tambe1, Mayur Pawar2, Krushna Parkhe3, Manas Badhan4, Prajakta Jagtap5
1Professor, Department of Computer Engineering, Sir Visvesvaraya Institute of Technology, Nashik, India
2BE Student, Department of Computer Engineering, Sir Visvesvaraya Institute of Technology, Nashik, India
3BE Student, Department of Computer Engineering, Sir Visvesvaraya Institute of Technology, Nashik, India
4BE Student, Department of Computer Engineering, Sir Visvesvaraya Institute of Technology, Nashik, India
5BE Student, Department of Computer Engineering, Sir Visvesvaraya Institute of Technology, Nashik, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The angle between body parts is crucial in the variety of asanas that yoga has to offer. If done correctly, yoga is an excellent form of physical exercise that is very good for your health. However, if yoga is practiced incorrectly, it can be harmful to one's health. Therefore, it's crucial to have a trainer when practicing yoga who can show you the proper form for each pose and keep an eye on it. This project carries a non-profit system that helps to strengthen the core muscles using yoga-like poses Virtual yoga asana practice is possible thanks to the totally accurate position detection provided by the suggested method. This system assists yoga enthusiasts with different yoga poses and validates them for correctness. Integrating computer vision techniques and deep learning techniques, the proposed system analyses the user’s human pose then based on the domain knowledge of yoga, the user is directed to correct the pose. Due to high computation requirements and a lack of available datasets, precise pose recognition in yoga is a challenging task. Different feature extraction and preprocessing techniques are applied to the dataset for the accurate detection of the yoga pose, achieving high accuracy just by using machine learning algorithms. The Human Pose Estimation technique, based on computer vision, is used to make the system effective and affordable.
Keywords: computer vision, feature extraction, learning (artificial intelligence), pose estimation
