Metraipolis › archive › dc_97ff5416c19b

Align3D: Image-Structure Co-Aligned 3D Scene Generation with Observability-Aware Fidelity-Preserving Guidance — arXiv:2611.08861

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.08861
License (as archived): CC0-1.0
Archived: 2026-11-09T22:45:07.541Z
SHA-256 of the archived text: 757b4ad15c1802d230d8cc0cc47ec5df688dc07e3f05eb4b7a2d0a5ae9845d7e
Stable URL of this page: https://metraipolis.com/documento/dc_97ff5416c19b

How to cite

Align3D: Image-Structure Co-Aligned 3D Scene Generation with Observability-Aware Fidelity-Preserving Guidance — arXiv:2611.08861 — arXiv (en), https://arxiv.org/abs/2611.08861 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_97ff5416c19b (Metraipolis archive dc_97ff5416c19b, sha256 757b4ad15c1802d230d8cc0cc47ec5df688dc07e3f05eb4b7a2d0a5ae9845d7e)

BibTeX

@misc{dc97ff5416c19b,
  title = {Align3D: Image-Structure Co-Aligned 3D Scene Generation with Observability-Aware Fidelity-Preserving Guidance — arXiv:2611.08861},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.08861},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_97ff5416c19b; sha256 757b4ad15c1802d230d8cc0cc47ec5df688dc07e3f05eb4b7a2d0a5ae9845d7e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.08861. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We present Align3D, an image-conditioned diffusion model that generates 3D indoor layouts that are both faithful to a single input photograph and physically plausible beyond the observed view. From the photograph we infer a confidence-calibrated scene graph and train a set-diffusion layout denoiser with image-structure co-alignment where confidence-weighted modality fusion balances image and scene-graph alignment using predicted confidences, while observability-aware fidelity-preserving guidance projects conflicting scene-quality gradients and reweighs them with an object-level confidence mask so under-observed degrees of freedom absorb physical priors without undoing the image-graph match. On 3D-FRONT, Align3D improves image-alignment and scene-quality metrics over recent baselines; once trained, it maps previously unseen photographs to layouts in a single feed-forward pass.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.08861 — 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 2026-11-09T22:37:47.206Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).