Study on the Application of IT-Supported Smart Classroom in Teaching Management

Students’ classroom learning status has a great impact on the formulation of teaching management measures and the improvement of classroom teaching quality, but the accuracy rate of most methods to capture students’ classroom status is only about 90% or lower. In order to improve the quality of teaching management, a smart classroom support technology based on information technology is designed. The structure of the traditional Visual Geometry Group-16 (VGG-16) model is improved by introducing local convolutional layers and reducing fully connected layers, thus improving the capability of the VGG. The improved VGG network is incorporated into the face detection model of the intelligent classroom. In the performance comparison experiments of the improved VGG, the precision rate of the improved VGG is 93.6%, which is higher than VGG (90.2%), convolutional Recurrent Neural Network (80.3%) and Convolutional Neural Network (76.5%). The result indicates that the capability of the improved VGG model is better than those of comparative models. Moreover, the improvement in the accuracy of the research method can support the higher teaching quality demand of the school’s smart classroom, further improve the analysis accuracy of teaching classroom behaviour and provide more reliable data support for teachers. Empirical analysis of the face detection model shows that the precision rate of the proposed face detection model is 88.4%, which is higher than the traditional face detection model. According to the methods and research results, it can be concluded that the improved VGG model based on convolutional neural network has a good effect when applied to classroom face detection algorithm. Significant contributions have been made in algorithm performance, algorithm optimisation and practical applications.