Constructing three-dimensional virtual spaces with the application of artificial intelligence

Abstract

This thesis examines the practical viability of generative Artificial Intelligence for producing VR-oriented 3D assets intended for cultural heritage visualization. It evaluates a five-stage workflow combining text-to-image generation, image-to-3D reconstruction using Tencent Hunyuan3D 2.0, and manual optimization in Blender. The workflow was applied to architectural and sculptural landmarks to test rendering performance on a Meta Quest 3 headset. Results indicate that while AI accelerates rapid ideation and initial mesh generation, substantial manual retopology and optimization remain critical bottlenecks. The study concludes that AI serves as a valuable assistive tool in the creative pipeline but does not yet replace expert manual workflows or measured heritage digitization.