Scratch PoseNet Exergame Prototyping for Learning Process Support in Physical Education

Abstract

This paper suggests the Exergaming approach of computer game control for a human motion in sport computer games in order to increase the level of physical activity of students in online classes, which takes into account the initial knowledge of the specialised courses of the IT curriculums. The approach is based on open access repository of computer programs in the Scratch-based block language (47 games for 28 summer sports, 16 games for 10 winter sports). Scratch programs for human motion recognition are based on the PoseNet neural network, which allows changing the principle of controlling game characters through the keyboard and mouse. In October 2023, ninety two first-year students majoring in Computer Science participated in a five-week online Exergaming as second part of the Physical Education course, among them 70 students (76%) successfully completed all stages. 57% of students felt motivated to continue using the Exergame approach and demonstrated the desire to achieve decent results, that indicates skill “Positive Thinking” in the discipline “Physical Education”, seven Hard Skills and four Soft Skills of Computing Curricula 2020 were indicated to compensate for the lack of classroom learning.