Student Classroom Teaching Behavior Recognition Based on DSCNN Model in Intelligent Campus Education

Abstract

In the contemporary era where artificial intelligence technology is becoming increasingly popular, its introduction into classroom teaching has become an important way to improve teaching quality. However, traditional methods for student behavior recognition suffer from low efficiency and insufficient accuracy. Therefore, a student classroom teaching behavior recognition scheme based on a dual stream convolutional neural network model was proposed in this study. The research focused on the visual geometry group and Res-Net method of convolutional neural networks and introduced knowledge distillation technology to optimize model efficiency. To further improve the performance of the model, an attention mechanism combined with a dual stream convolutional neural network model was ultimately constructed. The results confirmed that the recognition accuracy of the model reached 88.1% on the UCF-101 dataset and 89.4% on the STUDENT dataset. The accuracy rates of classroom teaching behavior recognition for students using mobile phones, writing, chatting, raising hands, and sleeping were 97.0%, 87.9%, 90.7%, 89.2%, and 96.1%, respectively. The processing speed of this model on the UCF-101 and STUDENT datasets was more than twice and 1.5 times that of traditional DSCNN models, respectively. Therefore, the proposed attention mechanism combined with the dual stream convolutional neural network model has demonstrated excellent recognition ability. This study provides key technical support for the intelligent transformation of the education industry.