Affiliations: School of Logistics Engineering, Wuhan University of Technology, China
Correspondence:
[*]
Corresponding authors. Meng Yu and Xiong Li, School of Logistics Engineering, Wuhan
University of Technology, China. E-mails: ymmona@126.com (Meng Yu) and jxlixiong@163.com (Xiong Li).
Abstract: Online Distance teaching for multiple smart classrooms by famous teachers, as an effective solver for the problem of lack of excellent teachers, has become a new popular teaching mode. However, one of the key problems to be solved urgently for this teaching mode is how to monitor children’s class status and effectively feedback their listening standing to teachers. Installation of intelligent pressure cushion on the chair of smart classroom to monitor children’s classroom state can be a powerful way to improve teaching effectiveness for the online distance teaching mode. This paper presents a new method for monitoring children’s classroom behavior based on intelligent cushion, which can identify basic children’s classroom behavior by classifying the original intelligent cushion pressure signal and evaluating the effectiveness of the classifier. To be concrete, the present method uses intelligent pressure cushion to collect data and denoises the original data by digital filter, and then extracts the time-domain and frequency-domain features of time-series pressure signals based on sliding time window. Finally, it uses machine learning to identify children’s status. In addition, by feature selection to reduce the data dimension, integrating different classifier to classify the extracted features, the efficiency of the present method is greatly improved.
