Metraipolis › archive › dc_e9f4bb3ec9c6
Image Clustering with Vision-Language Model Embeddings — arXiv:2610.33265
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.33265
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-27T07:59:39.317Z
- SHA-256 of the archived text
cc45aa4fab940e8fc4505fd60d7aec81940cddcab9c67786f36de544b54bb94e- Stable URL of this page
https://metraipolis.com/documento/dc_e9f4bb3ec9c6
How to cite
Image Clustering with Vision-Language Model Embeddings — arXiv:2610.33265 — arXiv (en), https://arxiv.org/abs/2610.33265 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e9f4bb3ec9c6 (Metraipolis archive dc_e9f4bb3ec9c6, sha256 cc45aa4fab940e8fc4505fd60d7aec81940cddcab9c67786f36de544b54bb94e)
BibTeX
@misc{dce9f4bb3ec9c6,
title = {Image Clustering with Vision-Language Model Embeddings — arXiv:2610.33265},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.33265},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e9f4bb3ec9c6; sha256 cc45aa4fab940e8fc4505fd60d7aec81940cddcab9c67786f36de544b54bb94e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.33265. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Pretrained vision-language models provide image representations whose suitability for clustering may vary with feature extraction, image domain, and partitioning algorithm. We compare nine representation pipelines across seven image benchmarks and four clustering configurations, yielding 252 baseline combinations. Eight pipelines extract visual features, including two vision-only references; the ninth generates descriptions and embeds the resulting text. With agglomerative clustering (Ward linkage) fixed, Qwen3-VL achieves the highest adjusted Rand index (ARI) on five datasets, including 0.925 on MNIST, while BLIP-2 leads on CIFAR-10 and CIFAR-100 with ARIs of 0.925 and 0.591, respectively. A separate sensitivity analysis compares raw and L2-normalized embeddings at their original dimensionality across all 63 dataset-representation pairs. Normalization preserves the highest-ARI representation in every dataset, increases ARI in 35 pairs, decreases it in 26, and leaves two numerically unchanged. On the SVHN dataset, the highest ARI remains low, rising from 0.097 to 0.098 after normalization. The evaluated caption-mediated pipeline obtains relatively high ARI on isolated digits but does not consistently exceed direct visual features. These findings support selecting and evaluating complete representation-clustering pipelines for the intended domain, with explicit attention to extraction and normalization choices. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.33265 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-27T07:55:30.814Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).