Construction of Sports Training Management Information System Using AI Action Recognition

• Main Authors: Cheng, D. (Author), Li, M. (Author), Wang, H. (Author)
• Format: Article
• Language: English
• Published: Hindawi Limited 2022
• Subjects: Action recognition; Convolutional neural network; Convolutional neural networks; Dance training; Development of science and technologies; Human being; Human thinking; Information management; Information use; Intelligent robots; Martial art; Neural network recognition; Recognition accuracy; Sports; Sports trainings
• Online Access: View Fulltext in Publisher

 With the development of science and technology, more and more fields have begun to use AI to provide convenient services for humans. Artificial intelligence (AI) refers to a new technology that uses human thinking to respond accordingly through computers and robots to assist human beings. Action recognition is an important research project that needs to be broken through in many industries, such as security system, martial arts instruction, and dance training. This paper aims to study a method for action recognition using AI technology and to build a sports training management information system. In this paper, a recognition model and related algorithms using a convolutional neural network (CNN) are proposed, and an intelligent sports training management information system is constructed. The system and the model are tested, the action recognition effect of 60 athletes in a university is tested, and the comparison with the traditional recognition algorithm is carried out. The results show that the CNN recognition accuracy test results used in this paper are generally more than 90%, while the traditional recognition accuracy rate is only about 75%, and the highest is not more than 86%; the training management information system of this paper takes about 15.7 s, and the maximum time is not more than 10 s, while the traditional recognition system takes about 15.7 s, which is about twice the time of the system in this paper. Therefore, it shows that the CNN recognition method in this paper has a significantly better effect on the recognition of athletes' movements, and the sports training management information system constructed in this paper is less time-consuming and faster and has certain feasibility. © 2022 Dali Cheng et al.