A dataset of spatial cognition experiment in virtual reality system

<p indent="0mm">Due to visual differences caused by projection, spatial cognition in virtual reality (VR) systems may deviate from that in the real world. Quantifying such deviation is an important research topic in spatial cognition science. However, the current lack of relevant experimental data, both domestically and internationally, limits the performance of artificial intelligence generative models in tasks in complex spatial scene understanding and reconstruction. This also impacts the immersive experience of applications such as 3D virtual scenes, game design, and the metaverse. To address this, the study developed a data acquisition system based on Google 3D Map, utilizing landmarks such as the Eiffel Tower and the Statue of Liberty as test scenarios, and conducted a series of “Cognition Testing of Spatial Scenes” experiments from September to December 2024. Through data processing methods such as coordinate transformation and outlier filtering, as well as quality control mechanisms including multiple verification and scoring systems, a dataset for spatial scene cognition was constructed. This dataset can be used to effectively measure deviations in human spatial cognition within scenarios. It not only contributes to the scientific understanding of human spatial cognition mechanisms but also provides scientific data support for the design and optimization of spatial scenes in VR systems.