This product uses a detection method based on the principle of computer vision, using the pytorch yolov5 dlib for deep learning to achieve student drowsiness monitoring. This system is deployed to the mobile terminal of strawberry pie, making it more convenient and portable for users. The early warning system for students' drowsiness and inattention behavior detection based on pytorch dlib makes full use of artificial intelligence in-depth learning. Once students have drowsiness or inattention behavior, the strawberry camera will detect and issue an alarm in time-voice prompt, alarm light warning, warning that students have entered a low-efficiency state, need to concentrate on the classroom, and automatically record relevant data at the same time for future reference and identification. Its function is to monitor and remind students of their own learning state, to give students active and intelligent learning protection, reduce students' inefficient state. Through the promotion of the use of this product, auxiliary teachers to solve the epidemic current students online class is not focused on the problem, promote the smooth completion of teaching arrangements, improve the efficiency of students' learning, is conducive to the smooth progress of national education, reduce the impact of the epidemic on the teaching tasks of major universities.