A Python Course Online and Offline Hybrid Classroom Management System Combining RN and SSD

Teaching is an important organic component of school education, and classroom management is a momentous means to realize the value of curriculum teaching and improve teaching quality. To establish a good teaching order and improve Python courses’ teaching quality, a hybrid online and offline classroom management system is constructed for Python courses by combining React Native and object detection models. This study first identifies student behavior in Python courses on the foundation of object detection models, and then constructs a hybrid online and offline classroom management system for Python courses on the foundation of the React Native framework. The results showed that compared with other models, the detection speed and average accuracy of the Single Shot MultiBox Detector model were the highest, with a detection speed of 53 and an average accuracy of 63.73%. When the number of iterations is 7000, the system has the highest accuracy in recognizing students’ listening behavior and playing mobile phone behavior, which are $\mathbf{0. 8 3 5}$ and 0.830, respectively. The accuracy of the system in recognizing students’ sleeping behavior and note taking behavior is 0.70 and 0.32, respectively. The accuracy and loss rate of the React Native framework for building systems for management are higher than those of the Cordova framework, with an accuracy rate of 98% and a loss rate of 18%. The system constructed through research can effectively improve the quality of Python course teaching and realize the value of course teaching.