Align3D: Image-Structure Co-Aligned 3D Scene Generation with Observability-Aware Fidelity-Preserving Guidance
Lists
Only you see this
Abstract
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.
BibTeX
@misc{zhou2026align,
author = {Chenliang Zhou and Yifan Wang and Andrzej Banburski-Fahey},
title = {{Align3D: Image-Structure Co-Aligned 3D Scene Generation with Observability-Aware Fidelity-Preserving Guidance}},
year = {2026},
eprint = {2611.08861},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.08861},
}Loading related content...