With the prosperity of virtual reality technology, especially head tracking technology, it has shown great potential in enhancing students' spatial cognitive abilities. This study aims to improve the performance of head tracking and apply it to the design of VR educational games to enhance students' immersion and spatial cognitive abilities.
The research proposed and optimized a Composite Fields for Human Pose Estimation (PIFPAF) algorithm. ShuffleNetv2 Block is introduced to enhance the efficacy and precision of feature extraction, and the impact of virtual reality educational games on students' spatial cognitive abilities is evaluated.
The experiment outcomes show that the improved algorithm performs excellently in head tracking accuracy, with a stable accuracy rate above 0.9. The average score for participants' immersion experience in virtual reality educational games is 3.55, skill challenge is 3.53 and feedback system is 3.70, demonstrating the positive role of virtual reality educational games in enhancing learners' cognitive skills. The outcomes of the analysis of variance indicate that immersion and skill challenge have a significant positive impact on spatial cognitive ability, with F values of 67.129 and 49.528, both p < 0.001.
The study not only validates the effectiveness of head tracking technology in enhancing the experience and learning outcomes of virtual reality educational games, but also provides empirical support for the design of future virtual reality educational games. The study emphasizes the importance of immersive experiences, skill challenges, and technological support in promoting the development of students' spatial cognitive abilities, which has significant practical implications for the field of education.
