Published Online:July 2025
Product Name:The IUP Journal of Computer Sciences
Product Type:Article
Product Code:IJCS020725
DOI:10.71329/IUPJCS/2025.19.3.21-33
Author Name:Adegbola Ebenezer Adebayo, Fatimah Binta Abdullahi and Francisca Nonyelum Ogwueleka
Availability:YES
Subject/Domain:Engineering
Download Format:PDF
Pages:21-33
This study presents the development and application of a facial emotion recognition (FER) system to enhance early childhood education (ECE) practices by accurately measuring the attention levels of students during classroom sessions. Traditional methods for gauging attention are often subjective and inconsistent, leading to a need for automated solutions that provide objective data to enhance teaching strategies. The primary objective was to design a robust FER model capable of detecting and quantifying attention levels to assist educators in refining their teaching methodologies. The study used machine learning (ML) algorithms, including decision tree (DT), random forest (RF) and convolutional neural networks (CNN), to build and evaluate the FER system. The empirical results demonstrated that the CNN model outperformed DT and RF models in accuracy, precision, recall, and F1-scores, achieving a high F1-score of 0.74 for classifying pupils’ attention levels. The study showed that CNN-based FER technology can effectively detect pupils’ emotional states and engagement levels, providing valuable insights for teachers to adjust their teaching methodologies. However, ethical considerations, such as privacy, data security, and algorithmic bias, must be addressed to ensure responsible implementation. The study recommends refining the FER model to enhance its accuracy across diverse demographics and expanding future research to explore the long-term impact of FER systems on pupils’ engagement and academic outcomes in educational settings.
Nonverbal communication is a term used to describe information exchanged without spoken words.