Deep Multi-Scale Residual Connected Neural Network Model for Intelligent Athlete Balance Control Ability Evaluation

Area of Science:

• Sports Science
• Biomechanical Engineering
• Artificial Intelligence

Background:

• Athlete balance control is crucial for sports performance and injury prevention.
• Current evaluation methods lack efficiency and intelligence, hindering athlete management.
• Demand for automated and intelligent balance assessment is growing in sports training.

Purpose of the Study:

• To propose a deep learning-based method for evaluating athlete balance control ability.
• To develop an end-to-end model that processes raw time-series pressure data for direct evaluation.
• To enhance practical utilization and efficiency in athlete balance assessment.

Main Methods:

• Utilizing time-series movement pressure measurement data.
• Implementing a multi-scale feature extraction scheme to capture diverse data patterns.
• Employing a residual connected neural network architecture for efficient deep learning model training.

Main Results:

• The proposed deep multi-scale residual connected neural network model demonstrated suitability for balance evaluation.
• Experimental validation on real athlete tests confirmed the model's effectiveness.
• The method showed promising results compared to existing approaches.

Conclusions:

• The developed deep learning model offers an accurate and efficient solution for athlete balance control evaluation.
• The end-to-end approach simplifies practical application in sports training and management.
• This intelligent evaluation method holds significant potential for real-world sports scenarios.