Recognition of Student Behavior through Actions in the Classroom.

Nowadays, scientific and technological advances are being applied in a variety of fields, in which education receives special attention. The students' learning status plays an important role in assessing the quality of the class, and several methods for identifying students' behaviors have been developed as a result. These methods are effective in monitoring the learning process and sending feedback back to the teacher through the classroom camera. Observing students, and providing feedback to the teacher so that the lesson can be quickly adjusted to the student's level of interest. Inheriting successes from existing methods and applying recent advances in computer vision. In this paper, the structures of actions and behaviors are proposed based on the functioning of students' body parts. These are the foundations for the extraction of necessary features of specific classroom behavior. Additionally, a database consisting of ten distinct actions has been constructed to facilitate the evaluation of the proposed method. This evaluation makes use of a cutting-edge deep learning model, allowing for accurate analysis and assessment of the identified actions. The experimental results show that the proposed method has achieved performance in detecting and classifying activities in real-time. [ABSTRACT FROM AUTHOR]