Lightweight online classroom expression monitoring method

Abstract

The invention provides a lightweight network classroom expression monitoring method, and the method comprises the following steps: acquiring facial pictures of students, and enabling different actioncombinations of eyes, mouths and eyebrows to correspond to different expressions to obtain an image training set; training a classroom expression monitoring model by using the image training set of the width learning network structure; shooting facial expressions of students in class to obtain static images of facial features; inputting the static image into a classroom expression monitoring modelto obtain an output value, and determining an expression mode to which the expression of the student belongs in the image. The expression recognition result is more effective, the expression mode recognition is more accurate, and the operation cost and the technical difficulty are low.

Description

technical field [0001] The invention relates to an expression monitoring method, which belongs to the field of image processing. Background technique [0002] Online classroom is a course form that has risen rapidly in recent years. It has been highly valued by governments, universities and enterprises around the world, and has become an important force to promote the reform of higher education. The online classroom uses the speed and convenience of video dissemination to realize the large-scale dissemination of the teaching process, and introduces interactive exercises to solve the problem of insufficient teaching feedback caused by one-way video dissemination. However, the teaching feedback ability provided by interactive exercises is still insufficient compared with the traditional offline teaching process. In offline courses, instructors can obtain feedback through students' facial expressions and questions from students, so as to make timely teaching adjustments, which...

AI Technical Summary

Problems solved by technology

However, the teaching feedback ability provided by interactive exercises is still insufficient compared with the traditional offline teaching process

In offline courses, instructors can obtain feedback through students' facial expressions and questions from students, so as to make timely teaching adjustments, which cannot be achieved in online classrooms

[0003] The neural network based on deep learning is a feasible direction for classroom expression monitoring, but with the promotion of online classrooms, there may be a large number of students in each classroom, so a deeper neural network is required to design a system with a higher recognition rate. network, which will bring about too much calculation, too long training time, and excessive memory consumption. On the other hand, the deeper deep structure network involves a large number of weights and parameters, which makes us have to continuously adjust the network. weights, parameters in order to achieve the best training results of the network

Images