Automatic Analysis System For Students Behavior In Online Classroom

Document Type : Primary Research paper


1 Bannari Amman Institute of Technology, Erode - 638401, India

2 CMR College of Engineering & Technology, Hyderabad - 501 401, India

3 Kongu Engineering College, Perundurai- 638060.


Student activity plays a crucial role in the learning process, either in the elearning
environment or in classroom. We use agitation, disruption, shifting pattern of head
posture, facial expressions and eye concentration to conclude meaningful information of
the student when engaged in an e-learning circumstances. Our method focuses on
recognising and estimating student activities during class time. In this paper introduced the
automatic analysis system and the computer vision techniques to track student’s classroom
activities. A multi-task cascade convolution neural network (MTCNN) automated analysis
system is proposed that is capable of monitoring student behaviours and performance. The
proposed automatic analysis system provides students with a high, medium and low degree
of attention during their learning environment. This system benefit both the teacher and the
students, so that teacher can trace the student’s interest in a specific subject. The proposed
automatic analysis system is evaluated to detect head movement, eye rotation and facial
perception and attention level of student focus in the classroom. Results shows that
proposed system has higher accuracy.


Volume 12, Issue 3 - Serial Number 3
ICMMNT-2021 International Virtual Conference on Materials, Manufacturing and Nanotechnology, 30th June, 2021.
June 2021
Pages 1032-1037