Metraipolis › archive › dc_bda74a549ee2

RoMa-$Ω$: What Feed-Forward 3D Models Know About Image Matching — arXiv:2609.09507

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.09507
License (as archived): CC0-1.0
Archived: 2026-10-16T10:59:26.083Z
SHA-256 of the archived text: e3d7c3ce32e4fcc73cb77cd09104fd43be6f4d5e26f83d52ab996d3556326cfb
Stable URL of this page: https://metraipolis.com/documento/dc_bda74a549ee2

How to cite

RoMa-$Ω$: What Feed-Forward 3D Models Know About Image Matching — arXiv:2609.09507 — arXiv (en), https://arxiv.org/abs/2609.09507 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bda74a549ee2 (Metraipolis archive dc_bda74a549ee2, sha256 e3d7c3ce32e4fcc73cb77cd09104fd43be6f4d5e26f83d52ab996d3556326cfb)

BibTeX

@misc{dcbda74a549ee2,
  title = {RoMa-$Ω$: What Feed-Forward 3D Models Know About Image Matching — arXiv:2609.09507},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.09507},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bda74a549ee2; sha256 e3d7c3ce32e4fcc73cb77cd09104fd43be6f4d5e26f83d52ab996d3556326cfb},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.09507. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Learned image matching has experienced significant progress in recent years, culminating in robust and accurate matchers such as RoMa, whose robustness is often attributed to its use of frozen DINO features. In a parallel development, feed-forward reconstruction models, such as VGGT, have been trained on ever-growing datasets to accurately regress dense 3D point maps and camera poses. The distinction between matchers and feed-forward reconstruction models has become increasingly blurred with the introduction of matching losses in models such as MASt3R and VGGT-$\Omega$. This raises a natural question: what do feed-forward 3D models know about image matching? In this work, we answer this question by analyzing three scenarios: (i) zero-shot matching of patch features, (ii) direct matching of 3D point predictions, and (iii) training a full matcher on top of the learned representations. We find that, despite performing poorly in zero-shot matching, especially in later layers, feed-forward reconstruction models provide strong representations for linear probing and full matching pipelines. We further show that, even without any training, their raw predictions alone enable competitive matching, albeit only under moderate viewpoint changes and modality gaps. Based on these insights, we retrain RoMa v2 by replacing its DINO backbone with VGGT-$\Omega$. Our resulting model, \ours, outperforms state-of-the-art matchers on a wide range of benchmarks, e.g. +8.1 mAA compared to RoMa v2 on WxBS.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-10 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.09507 — 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 2026-10-16T10:02:34.310Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).