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

2026-02-27

AI Summary

Existing neural implicit surface reconstruction methods struggle to recover thin structures and fine geometric details under noisy or unreliable geometric priors. To address this limitation, this work proposes GPU-SDF, a framework that explicitly models prior uncertainty through a self-supervised module and introduces an uncertainty-guided loss to preserve weak yet informative geometric cues. By integrating edge-aware signed distance fields with multi-view consistency constraints, the method effectively mitigates under-optimization in high-uncertainty regions. Extensive experiments demonstrate that GPU-SDF significantly enhances fine-grained geometric reconstruction quality in indoor scenes and can be seamlessly integrated as a plug-and-play component to augment existing neural reconstruction pipelines.

Technology Category

• Reasoning under Uncertainty: Uncertainty Representations
• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: 3D Computer Vision

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

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

Problem

• neural surface reconstruction
• geometric prior uncertainty
• fine detail recovery
• indoor scene reconstruction
• signed distance function

Innovation

• geometric prior uncertainty
• neural implicit surface reconstruction
• uncertainty-guided loss
• edge distance field
• multi-view consistency