Real-Time Classroom Behavior Analysis Using Deep Learning

Real-Time Classroom Behavior Analysis Using Deep Learning

Amulya G V
Harshwardhan G N
Supriya K P

Student of BE
Dept. of CSE

PES Institute of Technology & Management, Shivamogga-577204

Vidya B S

Student of BE

Sandeep K H

Assistant Professor

Dept. of CSE

PES Institute of Technology &

Management, Shivamogga-577204

Abstract - This article introduces a deep learning- based intelligent analysis model of the behaviour of teacher-student interactions in the classroom. In this model, the deep learning network YOLOv8 is used to detect typical teacher-student interactions once they have been encoded. It is useful for evaluating the behaviour of teacher-student interactions and classroom instruction. Student participation and interaction should be monitored to make the classroom instruction efficient. The AI-based methods of behaviour detection can be useful for monitoring students' attention and engagement. An intelligent real-time vision-based classroom has been included in the project for monitoring the students' academic performance and emotional states. Quality in classroom instruction will be improved if AI technology is integrated into education. The project aims at creating an intelligent evaluation system for the behaviour of teacher- student interactions in the classroom using the YOLO algorithm for object recognition and detection.

Key Words: Yolov8, deep learning.