Authors :
Tanmay Hande; Pranali Dhawas; Bhargavi Kakirwar; Aaditya Gupta
Volume/Issue :
Volume 8 - 2023, Issue 4 - April
Google Scholar :
https://bit.ly/3TmGbDi
Scribd :
https://t.ly/fFVMN
DOI :
https://doi.org/10.5281/zenodo.8281229
Abstract :
The COVID-19 epidemic has significantly
changed how we work out, with more people turning to
home fitness as a way to stay active during stay-at-home
orders. However, without access to professional trainers,
beginners may struggle to perform exercises with proper
form, increasing the risk of injury. Therefore, there is a
need for systems to monitor exercise performance for
both short- and long-term injury prevention. In this
study, we present an approach for accurately detecting
and correcting yoga postures using pose estimation
techniques with OpenCV and VGG-19 architectures
with GPU transfer learning. To precisely measure and
correct body posture during training sessions, the
suggested solution combines deep learning-based
algorithms and computer vision approaches. To confirm
the effectiveness of the VGG-19 model on the utilised
dataset, We conducted a large number of tests,
comparing the performance of several machine learning
and deep learning strategies for estimating yoga
postures. With a precision of 98.11 percent, the findings
show the usefulness of the suggested technique in
precisely recognising and correcting exercise postures.
The findings of this study have significant implications
for improving the effectiveness and safety of yoga
sessions and could be extended to other domains that
require precise human pose estimation.
The COVID-19 epidemic has significantly
changed how we work out, with more people turning to
home fitness as a way to stay active during stay-at-home
orders. However, without access to professional trainers,
beginners may struggle to perform exercises with proper
form, increasing the risk of injury. Therefore, there is a
need for systems to monitor exercise performance for
both short- and long-term injury prevention. In this
study, we present an approach for accurately detecting
and correcting yoga postures using pose estimation
techniques with OpenCV and VGG-19 architectures
with GPU transfer learning. To precisely measure and
correct body posture during training sessions, the
suggested solution combines deep learning-based
algorithms and computer vision approaches. To confirm
the effectiveness of the VGG-19 model on the utilised
dataset, We conducted a large number of tests,
comparing the performance of several machine learning
and deep learning strategies for estimating yoga
postures. With a precision of 98.11 percent, the findings
show the usefulness of the suggested technique in
precisely recognising and correcting exercise postures.
The findings of this study have significant implications
for improving the effectiveness and safety of yoga
sessions and could be extended to other domains that
require precise human pose estimation.
