IoT-Driven Community Sports Management: Machine Learning for Data Processing

抄録

Community sports management is influenced by complex social factors, and there are many sports events, making it difficult to achieve reliable sports identification and data processing. In order to improve the efficiency of community sports management, this paper proposes a two-person interactive action recognition algorithm based on weight fusion. The algorithm can recognize single-player atomic actions and two-player interactive actions. Among them, the single-player atomic action is recognized by the shapelet method, and the recognition results of two-player interactive actions are obtained by weight fusion of the recognition results of single-player atomic actions. In addition, based on the algorithm model, an intelligent management system for community sports is constructed. Finally, the analysis of the test results shows that the algorithm has higher accuracy in community sports recognition than existing algorithms. It also achieves higher practical performance and enables action recognition through terminal sensors, promoting community sports management.