Smart classroom is to realize the digitalization and intelligence of classroom teaching with the support of certain information technology and platforms, so that students’ learning develops in the direction of comprehensiveness and personalization. Based on artificial intelligence technology, this paper makes reasonable and effective planning for teaching resources and teaching process, and designs a set of smart classroom teaching mode covering before, during and after class. In the current study, the recognition of student classroom behavior technology plays an important role in objectively assessing the quality of classroom instruction. The SVM and IDT algorithms are used to classify the samples of students and training, respectively, and as a result, seven categories of behavior can be recognized, including raising hands, attentiveness to lecturing, writing, standing, reading, sleeping, and mobile phone usage. According to our study, after using the teaching model proposed above, the satisfaction level with AI-based learning resources was up to 78.6%, and the achievements of the students increased from 72.66 to 80.69. The difference between these groups of students was found to be significant compared with a control class with a traditional teaching model (P < 0.05).
