<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>With the development of big data, virtual course teaching has developed rapidly. In virtual teaching, the lecturer’s position is pivotal, and the virtual teacher’s lecture mainly conveys information through intuitive facial expressions, so the effectiveness of facial synthesis is crucial. However, traditional facial expression synthesis methods suffer from local distortion and lack of subtle expressions due to driving complex topological models and delays due to large computational effort. To this end, this paper proposes a facial expression extraction method based on feature point texture mapping. First, a video face expression feature point tracking algorithm is designed. Firstly, we design a video face expression feature point tracking algorithm. We use a monocular camera to capture facial expression images, use a texture mapping algorithm to detect faces and extract face feature points, and add a time threshold processing mechanism to the relevant filtering algorithm to detect face frame tracking to realize face feature point tracking in video. Secondly, the parameters are controlled by face expression animation, including head pose and face expression parameters. The head pose is solved by the Laplace algorithm, and the head rotation matrix and translation vector are output to establish the coordinate system. Finally, the stability and real-time performance of the feature point texture mapping algorithm are verified by expression capture test and recognition effect. The results of this paper show that: in the expression capture test, the average data transmission time is about 1ms, and the total expression transmission time of each video image frame is 32ms, which is 1.23ms and 9.8ms shorter than the traditional algorithm, respectively; in the expression recognition effect test, the successful recognition rate is increased by 10.2% on average compared with the traditional algorithm; and in the application to the classroom teaching effect test, 86.6% of the students are satisfied with the virtual teacher. 86.6% of the students highly agreed with the virtual teacher teaching, and 83.3% of the students thought that the model had a positive effect on the learners’ learning, which proved the superiority, real-time, stability and practicality of the algorithm designed in this paper compared with the traditional algorithm. The research work in this paper also provides solutions to the problems of virtual course teaching.</p>
</abstract>

With the development of big data, virtual course teaching has developed rapidly. In virtual teaching, the lecturer’s position is pivotal, and the virtual teacher’s lecture mainly conveys information through intuitive facial expressions, so the effectiveness of facial synthesis is crucial. However, traditional facial expression synthesis methods suffer from local distortion and lack of subtle expressions due to driving complex topological models and delays due to large computational effort. To this end, this paper proposes a facial expression extraction method based on feature point texture mapping. First, a video face expression feature point tracking algorithm is designed. Firstly, we design a video face expression feature point tracking algorithm. We use a monocular camera to capture facial expression images, use a texture mapping algorithm to detect faces and extract face feature points, and add a time threshold processing mechanism to the relevant filtering algorithm to detect face frame tracking to realize face feature point tracking in video. Secondly, the parameters are controlled by face expression animation, including head pose and face expression parameters. The head pose is solved by the Laplace algorithm, and the head rotation matrix and translation vector are output to establish the coordinate system. Finally, the stability and real-time performance of the feature point texture mapping algorithm are verified by expression capture test and recognition effect. The results of this paper show that: in the expression capture test, the average data transmission time is about 1ms, and the total expression transmission time of each video image frame is 32ms, which is 1.23ms and 9.8ms shorter than the traditional algorithm, respectively; in the expression recognition effect test, the successful recognition rate is increased by 10.2% on average compared with the traditional algorithm; and in the application to the classroom teaching effect test, 86.6% of the students are satisfied with the virtual teacher. 86.6% of the students highly agreed with the virtual teacher teaching, and 83.3% of the students thought that the model had a positive effect on the learners’ learning, which proved the superiority, real-time, stability and practicality of the algorithm designed in this paper compared with the traditional algorithm. The research work in this paper also provides solutions to the problems of virtual course teaching.

Design and implementation of a virtual teacher teaching system algorithm based on facial expression recognition in the era of big data

Zhengzhou Vocational University of Information and Technology

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

With the development of big data, virtual course teaching has developed rapidly. In virtual teaching, the lecturer’s position is pivotal, and the virtual...

Design and implementation of a virtual teacher teaching system algorithm based on facial expression recognition in the era of big data

Data publikacji: 17 lip 2023

Zakres stron: -

Otrzymano: 14 mar 2022

Przyjęty: 22 lis 2022

DOI: https://doi.org/10.2478/amns.2023.2.00053

Słowa kluczowe
Big data, Virtual teaching, Facial expression, Feature points, Texture mapping, Expression capture

© 2023 Rui Lu et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Design and implementation of a virtual teacher teaching system algorithm based on facial expression recognition in the era of big data

Data publikacji: 17 lip 2023

Zakres stron: -

Otrzymano: 14 mar 2022

Przyjęty: 22 lis 2022

DOI: https://doi.org/10.2478/amns.2023.2.00053

Słowa kluczoweBig data, Virtual teaching, Facial expression, Feature points, Texture mapping, Expression capture

© 2023 Rui Lu et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Słowa kluczowe
Big data, Virtual teaching, Facial expression, Feature points, Texture mapping, Expression capture