SMART HOME SPACE DESIGN INTEGRATING XR AND AIGC

Abstract

With the rapid advancement of extended reality (XR), artificial intelligence-driven generation (AIGC), and digital twin technologies, home space design is transitioning from "digital assistance" to a new phase of "intelligent symbiosis." Traditional virtual reality (VR) design faces limitations such as lengthy modeling cycles, conflicts between realism and real-time performance, and insufficient user engagement, making it inadequate for contemporary personalized and immersive design demands. This paper first analyzes the technological evolution of current home space design and highlights the necessity of integrating XR with AIGC; secondly, it establishes an intelligent design framework based on 3D Gaussian Spatter (3DGS) and generative spatial models (e.g., PanoWorld), enabling end-to-end generation of high-fidelity virtual scenes from user intent comprehension; thirdly, it employs an improved collaborative filtering algorithm to quantitatively evaluate spatial design preferences and similarity; finally, it compares simulation experiments with traditional design methods. Results demonstrate that the novel design approach achieves approximately 11% improvement in comprehensive metrics for spatial practicality, intuitiveness, and functional diversity compared to conventional methods, enhances user satisfaction by nearly 12%, and reduces the solution generation cycle from days to minutes. This study provides a novel technical pathway for advancing theoretical frameworks and engineering applications in smart home space design.