Yoga Pose Detection and Guidance Through an AI-Enhanced Chatbot

Information Science and Engineering

Yoga is a widely practiced form of exercise and meditation with numerous physical and mental health benefits. In recent years, the integration of technology into yoga practice has gained momentum, with applications ranging from virtual classes to personalized coaching. One such application is the automated detection and correction of yoga poses using machine learning techniques. This project focuses on developing a yoga pose detection system using transfer learning with the VGG19 convolutional neural network architecture. Transfer learning leverages the knowledge gained by pre-training models on large-scale datasets and adapts it to new tasks with smaller datasets, making it suitable for our pose classification problem. The VGG19 model, pre-trained on the ImageNet dataset, serves as a feature extractor for yoga pose images, capturing hierarchical visual representations. The workflow involves collecting a diverse dataset of yoga pose images, preprocessing them for uniformity, and annotating them with corresponding pose labels. We then fine-tune the pre-trained VGG19 model on the annotated dataset, replacing the final classification layers to suit our multi-class yoga pose classification task. Training the model involves optimizing hyperparameters and monitoring performance metrics to ensure effective learning. Evaluation of the trained model is performed using standard metrics such as accuracy, precision, recall, and F1 score on a separate validation or test dataset. The model's performance is assessed in terms of its ability to accurately classify various yoga poses under different conditions, including variations in pose, lighting, and background. The developed yoga pose detection system holds promise for assisting yoga practitioners and instructors in achieving correct pose alignment, providing real-time feedback, and tracking progress over time. Additionally, it serves as a foundation for further research and development in the intersection of machine learning and yoga practice, contributing to the advancement of technology-enhanced wellness solutions.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.