GCRayDiffusion: Pose-Free Surface Reconstruction via Geometric Consistent Ray Diffusion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Li-Heng, Zou, Zi-Xin, Liu, Chang, Jing, Tianjiao, Cao, Yan-Pei, Huang, Shi-Sheng, Fu, Hongbo, Huang, Hua
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908792706301952
author Chen, Li-Heng
Zou, Zi-Xin
Liu, Chang
Jing, Tianjiao
Cao, Yan-Pei
Huang, Shi-Sheng
Fu, Hongbo
Huang, Hua
author_facet Chen, Li-Heng
Zou, Zi-Xin
Liu, Chang
Jing, Tianjiao
Cao, Yan-Pei
Huang, Shi-Sheng
Fu, Hongbo
Huang, Hua
contents Accurate surface reconstruction from unposed images is crucial for efficient 3D object or scene creation. However, it remains challenging, particularly for the joint camera pose estimation. Previous approaches have achieved impressive pose-free surface reconstruction results in dense-view settings, but could easily fail for sparse-view scenarios without sufficient visual overlap. In this paper, we propose a new technique for pose-free surface reconstruction, which follows triplane-based signed distance field (SDF) learning but regularizes the learning by explicit points sampled from ray-based diffusion of camera pose estimation. Our key contribution is a novel Geometric Consistent Ray Diffusion model (GCRayDiffusion), where we represent camera poses as neural bundle rays and regress the distribution of noisy rays via a diffusion model. More importantly, we further condition the denoising process of RGRayDiffusion using the triplane-based SDF of the entire scene, which provides effective 3D consistent regularization to achieve multi-view consistent camera pose estimation. Finally, we incorporate RGRayDiffusion into the triplane-based SDF learning by introducing on-surface geometric regularization from the sampling points of the neural bundle rays, which leads to highly accurate pose-free surface reconstruction results even for sparse-view inputs. Extensive evaluations on public datasets show that our GCRayDiffusion achieves more accurate camera pose estimation than previous approaches, with geometrically more consistent surface reconstruction results, especially given sparse-view inputs.
format Preprint
id arxiv_https___arxiv_org_abs_2503_22349
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GCRayDiffusion: Pose-Free Surface Reconstruction via Geometric Consistent Ray Diffusion
Chen, Li-Heng
Zou, Zi-Xin
Liu, Chang
Jing, Tianjiao
Cao, Yan-Pei
Huang, Shi-Sheng
Fu, Hongbo
Huang, Hua
Computer Vision and Pattern Recognition
Accurate surface reconstruction from unposed images is crucial for efficient 3D object or scene creation. However, it remains challenging, particularly for the joint camera pose estimation. Previous approaches have achieved impressive pose-free surface reconstruction results in dense-view settings, but could easily fail for sparse-view scenarios without sufficient visual overlap. In this paper, we propose a new technique for pose-free surface reconstruction, which follows triplane-based signed distance field (SDF) learning but regularizes the learning by explicit points sampled from ray-based diffusion of camera pose estimation. Our key contribution is a novel Geometric Consistent Ray Diffusion model (GCRayDiffusion), where we represent camera poses as neural bundle rays and regress the distribution of noisy rays via a diffusion model. More importantly, we further condition the denoising process of RGRayDiffusion using the triplane-based SDF of the entire scene, which provides effective 3D consistent regularization to achieve multi-view consistent camera pose estimation. Finally, we incorporate RGRayDiffusion into the triplane-based SDF learning by introducing on-surface geometric regularization from the sampling points of the neural bundle rays, which leads to highly accurate pose-free surface reconstruction results even for sparse-view inputs. Extensive evaluations on public datasets show that our GCRayDiffusion achieves more accurate camera pose estimation than previous approaches, with geometrically more consistent surface reconstruction results, especially given sparse-view inputs.
title GCRayDiffusion: Pose-Free Surface Reconstruction via Geometric Consistent Ray Diffusion
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.22349