目前，大学生上课玩手机不再是个别现象，而授课老师在专注授课时又难以及时察觉和制止。对此，提出了一种基于卷积神经网络模型的学生听课状态应用。对拍摄获取的学生头像状态由网络模型自动识别并分析学生听课情况，低头族为疑似看手机对象或不在听课状态。课后，经由班主任进行针对性谈话了解情况并给予指导，以提高教学效果。

At present, playing with mobile phones in the class is no longer an individual phenomenon for college students, and most teachers cannot be aware and stop it in time due to the focus on teaching. This paper proposes the application of convolution neural network model in class status for students. The state of students obtained by shooting is automatically recognized by network model and the listening status is identified. After class, teachers can give some guidance to the students purposely, to improve the teaching effects.