Related Experiment Video
Updated: Sep 1, 2025
06:37
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
4.0K
A Deep Learning Method for Intelligent Analysis of Sports Training Postures
1Jiangxi Environmental Engineering Vocational College, Ganzhou 341000, China.
Computational Intelligence and Neuroscience
|August 12, 2022
Summary
This study introduces a new artificial intelligence method for analyzing sports training movements using a multiscale spatiotemporal graph convolution network. The enhanced model improves motion recognition accuracy and adapts to complex behavior changes in real-world conditions.
Area of Science:
- Computer Science
- Artificial Intelligence
- Sports Science
Background:
- Motion recognition is crucial for sports training posture analysis, but real-world factors like lighting and distance hinder accuracy.
- Existing models struggle to precisely capture human movement nuances due to environmental interferences.
Purpose of the Study:
- To develop an advanced motion training attitude analysis method robust to real-life interferences.
- To enhance the accuracy and adaptability of human movement recognition in sports training.
Main Methods:
- Constructing a spatiotemporal skeleton image and applying convolution operations.
- Utilizing a multiscale spatiotemporal graph convolution network for feature extraction.
- Implementing linear weighting and fusion of convolution results to capture varying action durations.
- Incorporating data augmentation techniques to handle information loss and increase dataset randomness.
Main Results:
- The proposed algorithm demonstrates significant improvements in model performance and recognition accuracy.
- The method effectively adapts to behavioral changes across different complexities.
- Enhanced ability to capture action characteristics of diverse time lengths.
Conclusions:
- The multiscale spatiotemporal graph convolution network offers a superior approach to motion recognition in sports training.
- The method provides a more accurate and adaptable solution for analyzing human movements under challenging real-world conditions.
