A classroom concentration model based on computer vision.

With the development of computer vision technology, automatic classroom behavior analysis has attracted more and more attention and the estimation of students' head pose is the most essential one of them. This paper proposed a classroom concentration model based on the head pose and position information collected by convolution neural networks. On the self-built dataset, the accuracy of the head pose classification model is 94.6%. In the real scene test, the results of our method are consistent with the manual analysis of the corresponding teaching videos. It means that the method can efficiently reflect real students' concentration situation during class and help teachers and experts carry out teaching analysis and rethink.