Leveraging Geometric Prior Uncertainty and Complementary Constraints for High-Fidelity Neural Indoor Surface Reconstruction

Qiyu Feng; Jiwei Shan; Shing Shin Cheng; Hesheng Wang

cs.CVFeb 27, 2026

Leveraging Geometric Prior Uncertainty and Complementary Constraints for High-Fidelity Neural Indoor Surface Reconstruction

Authors: Qiyu Feng, Jiwei Shan, Shing Shin Cheng, Hesheng Wang

Organizations: Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China. · Department of Mechanical and Automation Engineering and T Stone Robotics Institute, The Chinese University of Hong Kong, Hong Kong.

Abstract

Neural implicit surface reconstruction with signed distance function has made significant progress, but recovering fine details such as thin structures and complex geometries remains challenging due to unreliable or noisy geometric priors. Existing approaches rely on implicit uncertainty that arises during optimization to filter these priors, which is indirect and inefficient, and masking supervision in high-uncertainty regions further leads to under-constrained optimization. To address these issues, we propose GPU-SDF, a neural implicit framework for indoor surface reconstruction that leverages geometric prior uncertainty and complementary constraints. We introduce a self-supervised module that explicitly estimates prior uncertainty without auxiliary networks. Based on this estimation, we design an uncertainty-guided loss that modulates prior influence rather than discarding it, thereby retaining weak but informative cues. To address regions with high prior uncertainty, GPU-SDF further incorporates two complementary constraints: an edge distance field that strengthens boundary supervision and a multi-view consistency regularization that enforces geometric coherence. Extensive experiments confirm that GPU-SDF improves the reconstruction of fine details and serves as a plug-and-play enhancement for existing frameworks. Source code will be available at https://github.com/IRMVLab/GPU-SDF

Figures & tables

Fig. 1: (a) Monocular geometric priors for 3D reconstruction are often imprecise, especially for thin structures. (b) Comparison of different strategies to process geometric prior uncertainty. MonoSDF [ 4 ] uses priors directly, without handling uncertainty. DebSDF [ 7 ] relies on an implicit uncertainty that emerges during optimization, discarding supervision in unreliable regions and forcing reliance solely on RGB cues. In contrast, our method explicitly estimates prior uncertainty at the outset. We then employ an uncertainty-guided loss to modulate the influence of priors according to their reliability, rather than discarding them. Together with the additional geometric constraints we introduce, our method reconstructs fine-grained structures more effectively.

Fig. 2: Overview of the GPU-SDF framework. GPU-SDF reconstructs high-fidelity surfaces from multi-view RGB images and initial geometric priors (e.g., depth, normals). It consists of three parts: (1) Neural SDF pipeline: builds upon existing frameworks that represent the scene with an SDF and color field, and augments them by jointly learning an edge distance field as a robust geometric cue for fine structure reconstruction. (2) Prior uncertainty identification: a self-supervised module explicitly estimates confidence of geometric priors, enabling dynamic adjustment of their influence during optimization. (3) Loss functions: an uncertainty-guided consistency loss preserves weak but useful signals, while edge distance field supervision and multi-view regularization facilitate accurate recovery of thin and fine structures.

Fig. 3: Visualization of our self-supervised uncertainty estimation. (a) Uncertainty from horizontal flips, $U_{x}(D)$ ; (b) uncertainty from vertical flips, $U_{y}(D)$ ; (c) the combined uncertainty, $U(D)$ , which aligns closely with the ground-truth depth error map in (d).

Fig. 4: Illustration of the multi-view consistency regularization. For a pixel within a high-uncertainty region $P_{U}$ , its corresponding primary ray $r$ intersects the surface at point $s$ . A sphere with radius $H$ is centered at $s$ . Auxiliary rays, such as $r_{1}$ and $r_{2}$ , are then sampled, originating from points ( $m_{1}$ , $m_{2}$ ) on the sphere’s surface. If the first surface intersection of an auxiliary ray coincides with $s$ , its visibility flag is set to $v_{i}=1$ , and it is included in the consistency loss calculation. Otherwise, the flag is set to $v_{i}=0$ , and the ray is excluded.

Fig. 5: Visualization results of reconstructed mesh on ScanNet and Replica, with details highlighted in red boxes. The rightmost column shows an example of the prior uncertainty for the corresponding region. Note that the ground-truth mesh of the ScanNet dataset is generated from RGB-D sensors using voxel hashing, and therefore it is not accurate in fine-detail regions. These structures are highlighted with red dashed boxes and can be verified from the RGB images. Our method achieves finer geometric structures compared to previous methods.

Method	Prior	Acc $\downarrow$		Comp $\downarrow$		Chamfer $\downarrow$		F-score $\uparrow$
COLMAP	$\times$	4.7/	3.0	23.5/	9.5	14.1/	6.3	53.7/	65.8
Unisurf	$\times$	55.4/	4.5	16.4/	5.3	35.9/	4.9	26.7/	78.9
VolSDF	$\times$	41.4/	4.4	12.0/	8.3	26.7/	7.2	34.6/	69.5
GSRec	D+N	6.7/	4.1	6.9/	7.7	6.8/	5.8	59.9/	69.2
NeuRIS	N	5.0/	3.4	4.9/	7.0	5.0/	5.2	69.2/	75.4
MonoSDF	D+N	3.6/	2.7	3.9/	3.1	3.8/	2.9	77.1/	86.1

TABLE I: Quantitative performance on ScanNet/Replica. Bold indicates the best performance.

Methods	Priors	Metrics	Avg.	0e75f 3c4d9	036bc e3393	7f4d1 73c9c
VolSDF	$\times$	Acc $\downarrow$	6.8	5.3	5.3	9.7
		Chamfer $\downarrow$	7.6	6.5	7.6	8.6
		Prec $\uparrow$	44.2	39.5	47.5	45.7
B.A.	$\times$	Acc $\downarrow$	23.6	12.1	16.8	42.0
		Chamfer $\downarrow$	13.5	8.7	9.6	22.2
		Prec $\uparrow$	50.1	52.5	52.0	45.7

TABLE II: Quantitative Performance on ScanNet++. B.A. denotes Baked-Angelo. Bold indicates the best performance.

Method	Acc. $\downarrow$	Chamfer. $\downarrow$	Prec. $\uparrow$
ND-SDF	11.2	6.4	72.1
+D.U.(Horiz.)	10.8	6.3	73.1
+D.U.(Vert.)	10.6	6.2	72.8
+D.U.(Full)	10.7	6.2	73.3
+D.U.+N.U.	10.4	6.0	73.7
w/o EDF	11.0	6.4	72.7

TABLE III: Ablation Studies Results for key components of GPU-SDF.Bold indicates the best performance.

Fig. 6: Visualization results of ablation studies, with details highlighted in red boxes. Full Model achieves the finest detail in thin structures, demonstrating the effectiveness of our method.

UQ for 3D Reconstruction

See all 18 papers

CASA-SDF: Curriculum-Aware Spatial Adaptation with Curvature-Guided Density for Neural Implicit Surface Reconstruction
Jul 15, 2026Lei Yang, Weiqing Li, Zhiyong Su +1Signed Distance Functions 3D Reconstruction
School of Computer Science and Engineering, Nanjing University of Science and Technology, 200 Xiaolingwei, Nanjing, Jiangsu, China210094 · School of Automation, Nanjing University of Science and Technology, 200 Xiaolingwei, Nanjing, Jiangsu, China210094
NeuDonatello: Uncertainty-Aware Framework for Accurate Neural SDF Learning
Aug 27, 2026Alvin Jinsung Choi, Wanhee Kim, Taeyun Kim +33D Reconstruction Multi-View 3D Reconstruction
Department of Computer Science The University of Texas at Austin Texas, USA · School of Electrical Engineering Korea Advanced Institute of Science and Technology (KAIST) Daejeon, Republic of Korea · Robotics Program Korea Advanced Institute of Science and Technology (KAIST) Daejeon, Republic of Korea +2
First Shape, Then Meaning: Efficient Geometry and Semantics Learning for Indoor Reconstruction
May 5, 2026Remi Chierchia, Léo Lebrat, David Ahmedt-Aristizabal +3Signed Distance Functions 3D Reconstruction
School of Electrical Engineering & Robotics, Queensland University of Technology, Australia · Imaging and Computer Vision Group, CSIRO Data61, Australia

Jul 15, 2026J/K move · Enter open · S save

CASA-SDF: Curriculum-Aware Spatial Adaptation with Curvature-Guided Density for Neural Implicit Surface Reconstruction

Lei Yang, Weiqing Li, Zhiyong Su, Liang Xiao

School of Computer Science and Engineering, Nanjing University of Science and Technology, 200 Xiaolingwei, Nanjing, Jiangsu, China210094 · School of Automation, Nanjing University of Science and Technology, 200 Xiaolingwei, Nanjing, Jiangsu, China210094

Signed Distance Functions 3D Reconstruction Curriculum Learning Indoor Scene Reconstruction

Open paper AskPDF arXiv

Neural implicit representations have emerged as a powerful paradigm for 3D reconstruction. However, high-fidelity indoor surface reconstruction remains a significant challenge, primarily due to the pronounced geometric heterogeneity of indoor scenes. Large texture-less planar regions typically require stronger regularization to suppress high-frequency artifacts, while thin structures demand sharper, more adaptive representations to mitigate the spectral bias of multi-layer perceptrons (MLPs) and prevent over-smoothing. Existing approaches often rely on spatially indiscriminate prior supervision and a scene-global SDF-to-density transformation, which constrains their ability to balance planar smoothness and detail preservation. In this paper, we propose CASA-SDF (Curriculum-Aware Spatial Adaptation for SDF), a unified framework that addresses this challenge via complementary adaptations of supervision and representation capacity. Specifically, Hybrid Spatially-Adaptive Uncertainty Annealing (SAUA) fuses semantic and photometric uncertainties to construct a pixel-wise curriculum for monocular prior supervision. This strategy maintains regularization in reliable regions while attenuating unreliable supervision early in training to enable data-driven photometric refinement. Meanwhile, Curvature-Aware Locally Adaptive Density Transformation (CALADT) progressively modulates the sharpness of the SDF-to-density mapping via a curvature proxy to enhance the representation of thin structures. Extensive experiments on benchmark indoor datasets demonstrate that CASA-SDF improves surface completeness and detail recovery on high-frequency structures, without compromising the stability of planar surfaces.

Abstract

UQ for 3D Reconstruction

Explore similar work

CASA-SDF: Curriculum-Aware Spatial Adaptation with Curvature-Guided Density for Neural Implicit Surface Reconstruction

NeuDonatello: Uncertainty-Aware Framework for Accurate Neural SDF Learning

First Shape, Then Meaning: Efficient Geometry and Semantics Learning for Indoor Reconstruction