Metraipolis › archive › dc_780e528ee1f0

Two Global Crops Suffice: Locating Semantic Emergence in DINO-Style Self-Supervised Learning — arXiv:2609.28187

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.28187
License (as archived)
CC0-1.0
Archived
2026-10-08T12:35:15.868Z
SHA-256 of the archived text
ff12e56cbccd33349a15c9c5deb534c7e323feabd8c136502a6012c6e59fe1fe
Stable URL of this page
https://metraipolis.com/documento/dc_780e528ee1f0

How to cite

Two Global Crops Suffice: Locating Semantic Emergence in DINO-Style Self-Supervised Learning — arXiv:2609.28187 — arXiv (en), https://arxiv.org/abs/2609.28187 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_780e528ee1f0 (Metraipolis archive dc_780e528ee1f0, sha256 ff12e56cbccd33349a15c9c5deb534c7e323feabd8c136502a6012c6e59fe1fe)
BibTeX
@misc{dc780e528ee1f0,
  title = {Two Global Crops Suffice: Locating Semantic Emergence in DINO-Style Self-Supervised Learning — arXiv:2609.28187},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.28187},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_780e528ee1f0; sha256 ff12e56cbccd33349a15c9c5deb534c7e323feabd8c136502a6012c6e59fe1fe},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28187. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Self-supervised vision transformers trained with DINO-style objectives exhibit striking emergent semantic representation quality across visual tasks, yet the mechanisms underlying this behavior remain unclear. We present a systematic empirical dissection of the DINO family and show that semantic representations arise primarily from enforcing consistency between geometrically distinct global views of the same image instance. This instance-specific global alignment acts as the semantic anchor of DINO-style learning. Across controlled retraining experiments evaluated on semantic correspondence and a diverse suite of 2D and 3D downstream tasks, we find that patch-level masking objectives enhance semantics only when trained jointly with this global alignment, indicating that the iBOT objective refines and densifies existing semantic structure rather than creating it independently. In contrast, local-to-global view alignment does not substantially improve semantic qualities at fixed compute beyond a purely global alignment. Beyond training design, we revisit how semantic representation quality should be evaluated: while classification accuracy is the standard validation score, semantic correspondence provides a complementary axis that more reliably predicts downstream task performance. Together, these findings provide a functional decomposition of DINO-style learning and represent an important step toward understanding how semantic representations emerge in self-supervised vision models.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-24 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.28187 — 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 2026-10-08T12:12:35.780Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).