Project-based learning for machine learning in computer vision courses: Case study of Kazakhstan and Slovakia

Abstract

This study examines the effect of Project-Based Learning (PBL) on student motivation, engagement, and learning outcomes in the "Fundamentals of Machine Learning" course, focusing on computer vision applications. The research was conducted among third-year bachelor students in information technology programs. An experimental group, divided into four teams, implemented machine learning projects using TensorFlow, Keras, OpenCV, and DeepFace. Their results were compared with a control group following a traditional lecture-based approach. The experimental group showed a 60% increase in subject-specific motivation and a 55% rise in general learning motivation, significantly surpassing the control group. Post-test scores improved by 54% in the experimental group, contrasted with a 4% improvement in the control group, demonstrating that active, project-based activities boosted both theoretical and practical understanding of machine learning concepts. The results confirm that PBL fosters heightened enthusiasm for programming, deeper comprehension of machine learning models, and enhanced problem-solving skills in computer vision tasks. The study recommends broader adoption and further optimization of PBL approaches in technical education to increase student engagement, strengthen learning outcomes, and align coursework with real-world machine learning challenges.