Title: A method for identifying abnormal classroom behaviours of students based on multi-objective weighted learning

Authors: Lin Zou

Addresses: Teaching Quality Monitoring and Evaluation Centre, Tianjin Renai College, Tianjin, 301636, China

Abstract: In order to improve the accuracy of identifying abnormal behaviours among students and shorten the recognition time, a method based on multi-objective weight learning for identifying abnormal behaviours in student classrooms is proposed. Firstly, use mixed Gaussian background modelling to remove noise from student classroom monitoring images and improve image quality. Secondly, normalise the coordinates of student behaviour posture and extract classroom behaviour characteristics from both temporal and spatial features. Finally, taking student behaviour characteristics as input and student classroom abnormal behaviour recognition results as output, a multi-objective weight learning abnormal behaviour recognition model is constructed to obtain the recognition results of student classroom abnormal behaviour. The experimental results show that the method proposed in this paper can improve the recognition accuracy of abnormal classroom behaviour among students, with a recognition accuracy of 95.4%, and can shorten the recognition time, all of which are not less than 3.5 seconds.

Keywords: multi-objective weight learning; abnormal behaviour; student behaviour; classroom monitoring images.

DOI: 10.1504/IJBM.2025.143724

International Journal of Biometrics, 2025 Vol.17 No.1/2, pp.119 - 131

Received: 28 Dec 2023
Accepted: 22 Feb 2024
Published online: 06 Jan 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article