Space-adaptive artwork placement based on content similarities for curating thematic spaces in a virtual museum

Abstract

Virtual Reality (VR) provides curators with the tools to design immersive 3D exhibition spaces. However, manually positioning artworks in VR is labor-intensive, and most existing automated methods are limited in considering both artwork content and spatial characteristics, as well as accommodating curators’ design preferences. To address these challenges, this research presents a virtual exhibition authoring system that automatically generates optimized artwork placements for thematic space layouts, where exhibits are grouped by conceptual themes and arranged as clusters. First, the conceptual framework of thematic space layout is examined with fundamental elements and design process, and from this, a systematic and computational approach is devised. Our approach clusters artworks based on five content similarity factors—color, material, description, artist, and production date—and allows curators to adjust the importance of each factor and set limits on cluster sizes according to their design goals. A genetic optimization algorithm is employed to determine the placement of artworks, using four cost functions—intra-cluster distance, inter-cluster distance, intra-cluster intervisibility, and occupancy—to evaluate the arrangement with respect to spatial characteristics. Based on the three-phase research, it explores the opportunity of leveraging VR for curatorial planning by mapping desired themes adaptively to various exhibition spaces.