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

TL;DR

GPU-SDF is proposed, a neural implicit framework for indoor surface reconstruction that leverages geometric prior uncertainty and complementary constraints that improves the reconstruction of fine details and serves as a plug-and-play enhancement for existing frameworks.

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 (6)

• Figure 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 monosdf uses priors directly, without handling uncertainty. DebSDF DebSDF 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.
• Figure 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.
• Figure 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).
• Figure 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.
• Figure 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.