SDXL model-based optimization for interior design: Data-driven and deep learning methods
PLOS One · 04.02.2026 · 21 (2) · e0342258
DOI: 10.1371/journal.pone.0342258 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 04.02.2026. Verfasst von Xiaofei Zhou und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Art and Design, Dalian Art College, Dalian, Liaoning Province, China. Hauptthemen laut MEDLINE: Deep Learning, Models, Theoretical. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: 2025 Planning Research Topics of the Association of Employment for University Graduates, Liaoning Provincial Department of Education, Liaoning Cultural and Creative Industry Collaborative Innovation Research Center.
Abstract (Originalsprache)
This study proposes a novel, domain-specific optimization framework for the Stable Diffusion XL (SDXL) model, addressing the critical challenges of structural consistency and aesthetic fidelity in AI-assisted interior design. Unlike generic applications of diffusion models, this research introduces a systematic pipeline integrating automated semantic cleaning with a rigorous hyperparameter optimization strategy. A high-quality, annotated dataset was constructed using a semi-automated YOLO-based filtering process to minimize noise. Furthermore, we established an empirically validated training protocol—combining optimal Dropout rates, L1/L2 regularization, and dynamic learning rates—specifically tuned to preserve the geometric constraints of interior spaces. Experimental results demonstrate that this optimized framework significantly outperforms baseline models, achieving superior Fréchet Inception Distance (FID), Structural Similarity Index (SSIM), and Learned Perceptual Image Patch Similarity (LPIPS) scores, alongside robust CLIP Semantic Alignment. Furthermore, a systematic ablation study confirms that while domain-specific data provides the foundation, our semantic cleaning pipeline and structural regularization are critical for achieving high geometric fidelity, reducing FID by 51.1% compared to the baseline. The study contributes a technically robust methodology for adapting large-scale diffusion models to the specialized requirements of spatial design.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- School of Art and Design, Dalian Art College, Dalian, Liaoning Province, China.
- Department of Interior Architecture, Daegu University, Gyeongsan City, North Gyeongsang Province, South Korea.
- School of Architecture and Fine Art, Dalian University of Technology, Dalian, Liaoning Province, China.
ORCID-Profile: Xiaofei Zhou
Förderung
- 2025 Planning Research Topics of the Association of Employment for University Graduates
- Liaoning Provincial Department of Education
- Liaoning Cultural and Creative Industry Collaborative Innovation Research Center
Zitieren
APA
Zhou, X., Kim, S., & Chen, Y. (2026). SDXL model-based optimization for interior design: Data-driven and deep learning methods. PLOS One, 21(2), e0342258. https://doi.org/10.1371/journal.pone.0342258
Vancouver
Zhou X, Kim S, Chen Y. SDXL model-based optimization for interior design: Data-driven and deep learning methods. PLOS One. 2026;21(2):e0342258. doi:10.1371/journal.pone.0342258
BibTeX
@article{Zhou2026,
author = {Zhou, Xiaofei and Kim, Soohong and Chen, Yan},
title = {SDXL model-based optimization for interior design: Data-driven and deep learning methods},
journal = {PLOS One},
year = {2026},
volume = {21},
number = {2},
pages = {e0342258},
doi = {10.1371/journal.pone.0342258}
}
EUVIMED