Date of Publication :25th July 2024

Abstract:This research introduces a pioneering web-based platform that melds ancient wellness practices with cutting- edge technology, transforming the realms of yoga and fitness. Utilizing ml5.js and sophisticated machine learning, the platform refines exercise performance monitoring, focusing on the revital- ization of Surya Namaskar—a foundational yoga sequence—and various exercises such as squats, push-ups, and lunges.Surya Namaskar, or sun salutation, offers significant physical and men- tal benefits, including enhanced strength, flexibility, endurance, cardiovascular health, mindfulness, and overall vitality. Our platform employs machine learning to provide supervised pose recognition, ensuring precision and fostering disciplined, goal- oriented fitness routines.The project encompasses three main sections: Surya Namaskar Posture Detection, Exercise Detection, and Yoga Posture Detection. Users receive real-time feedback and performance metrics, optimizing their fitness journeys. p5.js creates the user interface and captures pose lines, while the pre- trained ml5.js model, based on PoseNet, ensures accurate pose recognition.Acknowledging existing research in deep learning and pose estimation, our project addresses limitations through seamless web integration and customized neural networks for superior detection efficiency. Methodology includes data collec- tion, model training, and real-time pose classification, empha- sizing user interaction, pose correction, and exercise tracking.To synchronize continuous feature extraction and model detection in dynamic sequences like Surya Namaskar, we employ a mutex lock mechanism, ensuring accurate and stable model transitions. Comprehensive user performance reporting includes metrics such as calories expended.In essence, this research offers a holistic approach to wellness, elegantly combining traditional yoga practices with modern technology to enhance physical and mental well-being.

Reference :