AI Enhanced Yoga Pose Detection And Alignment Using Deep Learning
V. S. Lavanya, C N A Kalila, Keshav Aditya R. Premkumar, R Rohith · 2024
Yoga practice offers numerous health benefits, but incorrect poses can lead to injuries and hinder progress. This project leverages the power of deep learning, specifically Convolutional Neural Networks (CNNs) and TensorFlow Movenet, to address this challenge. We introduce an AI-enhanced system that performs real-time yoga pose detection and alignment. Users can choose from a variety of yoga poses, access step-by-step instructions, and activate the system. When the correct pose is achieved, 17 key body points turn green, activating a timer with an alarm-like sound. If the pose becomes incorrect, the timer and sound cease, providing immediate feedback. This project aims to enhance yoga practice safety, provide real-time guidance, and track progress, fostering a more effective and enjoyable yoga experience.