A Comprehensive Survey on AI-Based Student Behavior Recognition in Classrooms: Datasets, Models, and Practi-Cal Challenges
摘要
Detecting student behavior using Artificial Intelligence (AI) is an important research direction aimed to improve classroom management and personaliz-ing education. This paper provides an overview of term, definition, method, datasets, and edge models in this field. We analyze image-based, audio-based, and multimodal approaches, and evaluate commonly used datasets. In addition, the paper highlights practical challenges related to data, models, ethics, and privacy, and proposes future development directions, emphasizing the need for Vietnam-specific datasets and the responsible application of AI.