The Capture and Evaluation System of Student Actions in Physical Education Classroom Based on Deep Learning

Socially, politically, and morally, the world of sport is still changing. On the other hand, technology has been the most prevalent transition in the sport over the last century. Thanks to modern science, athletes can now go higher, run quicker, and, most importantly, remain healthy. Although academics, agencies, and policymakers had already urged physical education teachers to use technology in their classrooms, in many of these situations, technology is used for administrative purposes, including tracking enrolment and measuring, documenting, and reporting students’ work. Thus, this paper suggests an intelligent Student Actions Evaluation System using Deep Learning (iSAES-DL)for student monitoring in physical education. This model uses the deep convolution neural network for the classification of risky actions. This model further evaluates the learners’ degree of learning, retention, and achievements and suggests improvements and corrective measures. It highlights the benefits, uses, and limitations of applying deep learning techniques and IoT devices to develop learning analytics systems in the physical, educational domain. Eventually, output criteria such as comprehension, concentration, retention, and learner attainment are given a feature-by-feature analysis of the proposed methodology and traditional teaching-learning approaches. Finally, the classification algorithm is contrasted to other deep learning algorithms with an F1-score of 97.86%.