Engagement Detection for Offline Education Based on Computer Vision
In order to detect the overall engagement level of the class during classroom learning,a computer vision-based classroom engage-ment detection system is designed.The head deflection angle is calculated based on the location of key points on the face,which is used to de-termine whether the line of sight is located in the engagement area.The ViViT model is adopted to perform four classifications of learners' en-gagement.Face detection information,line of sight information,and engagement classification information are finally combined to calculate the overall engagement of the class.The experimental results prove that the system can quickly and effectively evaluate the overall engagement of the class at a certain moment based on the monitoring video.It can provide teachers with objective and accurate engagement data,which can help teachers adjust teaching methods in a targeted and timely manner to improve teaching quality.
国家科技期刊平台
NSTL
万方数据
