EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Unsupervised Hierarchical Visual Taxonomy of Marble Natural Stone Using Cluster-Aware Self-Supervised Vision Transformers

Margarida Figueiredo, Carlos M. A. Diogo, Gustavo Paneiro, Pedro Amaral, António Alves de Campos

Applied Sciences · 23.04.2026 · 16 (9) · 4137

DOI: 10.3390/app16094137 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 23.04.2026. Verfasst von Margarida Figueiredo und 4 weiteren Autor:innen. Die Arbeit nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: PRR − Recovery and Resilience Plan of the European Union.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The marble industry relies on proprietary commercial names rather than objective visual categories, creating market inefficiencies for stakeholders who select stones based on appearance. Supervised classification perpetuates this problem by replicating inconsistent commercial labels instead of discovering intrinsic visual structure. We propose an unsupervised pipeline combining a two-stage training strategy: A pure self-supervised pretraining followed by cluster-aware fine-tuning of a DINO Vision Transformer, with empirically selected dimensionality reduction and agglomerative hierarchical clustering. Systematic ablation studies on 1480 marble images spanning 10 commercial varieties validate each design choice: cluster-aware training at k = 10 yields geometrically improved embeddings over the self-supervised baseline (mean Silhouette Score 0.693 ± 0.053 vs. 0.660 ± 0.030; mean Davies–Bouldin Index 0.386 ± 0.075 vs. 0.569 ± 0.012; N = 9 independent evaluations across 3 data partitions × 3 training initializations). The resulting taxonomy reveals three phenomena invisible to commercial classification: cross-category merging of visually indistinguishable stones carrying different market names, intra-category splitting of heterogeneous sub-populations within single varieties, and coherent grouping where commercial and visual boundaries coincide, with all three confirmed in every independent run. We further demonstrate that standard extrinsic metrics are misaligned with unsupervised taxonomy objectives when reference labels encode the inconsistencies the method aims to resolve. Validating this methodology across diverse stone types, larger datasets, and varied acquisition conditions represents a natural and necessary next step toward establishing its cross-domain generalizability.

Förderung

  • PRR − Recovery and Resilience Plan of the European Union

Zitieren

APA

Figueiredo, M., Diogo, C. M. A., Paneiro, G., Amaral, P., & Campos, A. A. D. (2026). Unsupervised Hierarchical Visual Taxonomy of Marble Natural Stone Using Cluster-Aware Self-Supervised Vision Transformers. Applied Sciences, 16(9), 4137. https://doi.org/10.3390/app16094137

Vancouver

Figueiredo M, Diogo CMA, Paneiro G, Amaral P, Campos AAD. Unsupervised Hierarchical Visual Taxonomy of Marble Natural Stone Using Cluster-Aware Self-Supervised Vision Transformers. Applied Sciences. 2026;16(9):4137. doi:10.3390/app16094137

BibTeX
@article{Figueiredo2026,
  author = {Figueiredo, Margarida and Diogo, Carlos M. A. and Paneiro, Gustavo and Amaral, Pedro and Campos, António Alves de},
  title = {Unsupervised Hierarchical Visual Taxonomy of Marble Natural Stone Using Cluster-Aware Self-Supervised Vision Transformers},
  journal = {Applied Sciences},
  year = {2026},
  volume = {16},
  number = {9},
  pages = {4137},
  doi = {10.3390/app16094137}
}
Kostenlos starten →