Design of Inquiry-Based Teaching Experiments for AI Courses Based on Optimized YOLO11n Model: A Case Study of Classroom Behavior Recognition

Targeting the prominent teaching pain point of overemphasizing theory while neglecting practice in university artificial intelligence courses and adhering to the talent cultivation requirements of Emerging Engineering Education (EEE), this paper proposes and implements an inquiry-based teaching experiment scheme based on the optimized YOLO11n model, with classroom behavior recognition as the application carrier. Guided by real engineering problems, the experiment leads students to complete the full-cycle engineering practice covering demand analysis, model optimization, dataset construction and performance evaluation, and promotes the organic integration of algorithm theory and teaching practice. In the teaching implementation, this paper adopts the teaching mode of “problem orientation - project-driven learning - team collaboration”, and establishes a diversified assessment mechanism combining process evaluation and outcome evaluation. Teaching practice shows that the proposed scheme effectively stimulated students' learning initiative, significantly improved their engineering practice ability and teamwork literacy, helped students complete multiple innovative practice projects, and provided an exemplary implementation scheme for the practical teaching reform of artificial intelligence courses.