Analysis on the Steps of Physical Education Teaching Based on Deep Learning

Abstract

The rapid progress of the internet of things and artificial intelligence has brought new opportunities for the construction and development of intelligent sports. This paper designs an analysis and evaluation system of physical education teaching steps based on deep learning technology. The intelligent wearable devices are used to conduct real-time dynamic monitoring of students' exercise steps and heart rate in class so as to build a sports teaching activity data set. The authors analyze the time step sequence based on transformer deep model to realize the estimation of motion effect. In addition, they propose a hierarchical fusion model based on transformer, which makes full use of the steps and heart rate information to predict the abnormal situation in physical education. The experimental results show the effectiveness of the system.