基于MMAction2模型的体育课堂教学行为评价系统设计与应用

打开文本图片集

Abstract：Leveraging the advanced MMAction2 model，the study applies it forthe first time to the evaluation of teaching behavior in physical education clasroom， and designs a complete system encompassing multi-inteligence algorithm analysisand visual feedback functions，soas to achieve automation and intellgence from data collection to behavior analysis.This research shows that this system consists of four core modules such as the perception layer （responsible for data collection and input），the platform layer （for data processing and storage），the model layer （for behavior recognition and analysis），and the application layer （providing data visualization and result feedback），and these modules efficiently collaborate to construct a comprehensive evaluation system for teaching behavior. Andpractical tests demonstrate that this system achieves a 92% accuracy rate in behavior recognition and a 95% consistency between analysis results and manual annotations in basketballclassroom teaching， which significantly enhances the eficiency and accuracy of teaching and evaluation.

Keywords： teaching behavior in physical education classroom； artificial inteligence；evaluation system； neural network;MMAction2 model

2023年9月，习近平总书记提出“新质生产力”发展理念，倡导以前沿技术催生新产业、新模式、新动能，推动产业创新。（剩余15958字）