Semantic gaussian splatting-enhanced facility management within the framework of IFC-graph

Liu, Jiucai, Li, Haijiang ORCID: https://orcid.org/0000-0001-6326-8133 and Khudhair, Ali ORCID: https://orcid.org/0000-0002-5062-7448 2025. Semantic gaussian splatting-enhanced facility management within the framework of IFC-graph. Presented at: EG-ICE 2025: International Workshop on Intelligent Computing in Engineering, Glasgow, UK, 1-3 July 2025. Published in: Moreno-Rangel, A. and Kumar, B. eds. EG-ICE 2025. Glasgow: University of Strathclyde Publishing,

Preview

PDF - Published Version Download (1MB) | Preview

Abstract

The indoor environment of as-built structures undergo modifications over time, deviating from their original design and complicating facility management. Existing methods struggle to produce high-fidelity 3D models to represent that change. Gaussian splatting, a recent reconstruction technique, enables real-time rendering with high visual quality, offering a promising solution. This paper proposes a method that generates semantic Gaussian splats from panoramic images and integrates them into an IFC-graph for facility management. The approach enhances facility recognition and retrieval while enabling high-fidelity rendering of indoor environments. A case study demonstrates two key applications: semantic-guided facility retrieval and high-fidelity rendering, improving the semantic richness and visual quality of as-built BIM models.

Item Type:	Conference or Workshop Item (Paper)
Date Type:	Publication
Status:	Published
Schools:	Schools > Engineering
Publisher:	University of Strathclyde Publishing
ISBN:	9781914241826
Related URLs:	https://doi.org/10.17868/strath.00093259
Date of First Compliant Deposit:	2 July 2025
Date of Acceptance:	9 June 2025
Last Modified:	07 Jul 2025 09:30
URI:	https://orca.cardiff.ac.uk/id/eprint/178376

Actions (repository staff only)

Edit Item