Light Transport-aware Diffusion Posterior Sampling for Single-View Reconstruction of 3D Volumes

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Leonard, Ludwic, Thuerey, Nils, Westermann, Ruediger
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912298458677248
author Leonard, Ludwic
Thuerey, Nils
Westermann, Ruediger
author_facet Leonard, Ludwic
Thuerey, Nils
Westermann, Ruediger
contents We introduce a single-view reconstruction technique of volumetric fields in which multiple light scattering effects are omnipresent, such as in clouds. We model the unknown distribution of volumetric fields using an unconditional diffusion model trained on a novel benchmark dataset comprising 1,000 synthetically simulated volumetric density fields. The neural diffusion model is trained on the latent codes of a novel, diffusion-friendly, monoplanar representation. The generative model is used to incorporate a tailored parametric diffusion posterior sampling technique into different reconstruction tasks. A physically-based differentiable volume renderer is employed to provide gradients with respect to light transport in the latent space. This stands in contrast to classic NeRF approaches and makes the reconstructions better aligned with observed data. Through various experiments, we demonstrate single-view reconstruction of volumetric clouds at a previously unattainable quality.
format Preprint
id arxiv_https___arxiv_org_abs_2501_05226
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Light Transport-aware Diffusion Posterior Sampling for Single-View Reconstruction of 3D Volumes
Leonard, Ludwic
Thuerey, Nils
Westermann, Ruediger
Computer Vision and Pattern Recognition
Machine Learning
I.2.10; I.4.5
We introduce a single-view reconstruction technique of volumetric fields in which multiple light scattering effects are omnipresent, such as in clouds. We model the unknown distribution of volumetric fields using an unconditional diffusion model trained on a novel benchmark dataset comprising 1,000 synthetically simulated volumetric density fields. The neural diffusion model is trained on the latent codes of a novel, diffusion-friendly, monoplanar representation. The generative model is used to incorporate a tailored parametric diffusion posterior sampling technique into different reconstruction tasks. A physically-based differentiable volume renderer is employed to provide gradients with respect to light transport in the latent space. This stands in contrast to classic NeRF approaches and makes the reconstructions better aligned with observed data. Through various experiments, we demonstrate single-view reconstruction of volumetric clouds at a previously unattainable quality.
title Light Transport-aware Diffusion Posterior Sampling for Single-View Reconstruction of 3D Volumes
topic Computer Vision and Pattern Recognition
Machine Learning
I.2.10; I.4.5
url https://arxiv.org/abs/2501.05226