Related Experiment Video
Updated: Jan 7, 2026
07:56
Resting-State Connectivity and Neuroimaging of Prefrontal Cortex Activity During a Block-Design Yoga Asana Practice Using fNIRS
Published on: June 24, 2025
758
Automated identification of Shankha Prakshalana yoga poses with machine learning techniques
Abhishek Sharma1, Vaidehi Sharma2, Premansh Sharma3
1Electronics and Communication Engineering, LNMIIT, Jaipur, Rajasthan, India. abhisheksharma@lnmiit.ac.in.
Scientific Reports
|December 31, 2025
Summary
This study introduces an automated yoga pose detection system for Shankha Prakshalana. The Random Forest model achieved 99.6% accuracy, advancing computer vision in yoga practice.
Area of Science:
- Computer Vision
- Machine Learning
- Yoga Studies
Background:
- Traditional yoga practices like Shankha Prakshalana lack objective monitoring tools.
- Integrating technology can enhance the accessibility and efficacy of yoga training.
Purpose of the Study:
- To develop an automated system for detecting the yoga pose Shankha Prakshalana.
- To evaluate machine learning models for accurate pose classification from video data.
Main Methods:
- Utilized computer vision techniques for feature extraction from yoga videos.
- Trained and evaluated supervised and unsupervised machine learning algorithms.
- Developed a dataset of annotated Shankha Prakshalana practice videos.
Main Results:
- The Random Forest classifier achieved a 99.6% recognition rate.
- Demonstrated the effectiveness of machine learning in automated yoga pose detection.
- Identified optimal ML architectures for pose classification.
Conclusions:
- The developed system offers a novel approach to monitoring and improving yoga practice.
- This research bridges traditional yogic techniques with modern computer vision technology.
- Potential applications include enhanced yoga training and practice analysis.
