Advancing Structured Priors for Sparse-Voxel Surface Reconstruction

2026-01-25

AI Summary

This work proposes a novel surface reconstruction method that overcomes the limitations of existing sparse voxel-based and 3D Gaussian splatting approaches. While conventional sparse voxel methods suffer from slow convergence due to uniform dense initialization and poor exploitation of scene structure, 3D Gaussian splatting achieves rapid convergence but lacks geometric fidelity. The proposed method integrates the strengths of both paradigms by introducing geometry-prior-guided structured voxel initialization, combined with multi-view depth-based geometric supervision and sparse voxel rasterization. This enables efficient per-scene optimization that simultaneously achieves high-quality reconstruction, significantly improving geometric accuracy, surface completeness, and fine structural detail recovery while maintaining fast convergence. Extensive evaluations on standard benchmarks demonstrate superior performance over current state-of-the-art methods.

Technology Category

• Computer Vision: 3D Computer Vision
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Reconstructing accurate surfaces with radiance fields has progressed rapidly, yet two promising explicit representations, 3D Gaussian Splatting and sparse-voxel rasterization, exhibit complementary strengths and weaknesses. 3D Gaussian Splatting converges quickly and carries useful geometric priors, but surface fidelity is limited by its point-like parameterization. Sparse-voxel rasterization provides continuous opacity fields and crisp geometry, but its typical uniform dense-grid initialization slows convergence and underutilizes scene structure. We combine the advantages of both by introducing a voxel initialization method that places voxels at plausible locations and with appropriate levels of detail, yielding a strong starting point for per-scene optimization. To further enhance depth consistency without blurring edges, we propose refined depth geometry supervision that converts multi-view cues into direct per-ray depth regularization. Experiments on standard benchmarks demonstrate improvements over prior methods in geometric accuracy, better fine-structure recovery, and more complete surfaces, while maintaining fast convergence.

Problem

• sparse-voxel surface reconstruction
• 3D Gaussian Splatting
• sparse-voxel rasterization
• geometric priors
• surface fidelity

Innovation

• structured priors
• sparse-voxel initialization
• depth geometry supervision
• 3D Gaussian Splatting
• surface reconstruction