PAGaS: Pixel-Aligned 1DoF Gaussian Splatting for Depth Refinement

2026-04-23

AI Summary

This work addresses the limited geometric accuracy of depth estimation in multi-view stereo by proposing a one-degree-of-freedom Gaussian splatting representation with pixel-aligned constraints. The method strictly confines Gaussian modeling within the back-projected pixel volume, treating depth as the sole optimization variable. By tightly integrating geometric priors with learning-based mechanisms, it achieves high-fidelity surface reconstruction while maintaining computational efficiency. Extensive experiments demonstrate that the proposed approach outperforms both geometric and learning-based state-of-the-art multi-view stereo methods across multiple 3D reconstruction benchmarks, producing depth maps with superior geometric fidelity and enhanced detail preservation.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Gaussian Splatting (GS) has emerged as an efficient approach for high-quality novel view synthesis. While early GS variants struggled to accurately model the scene's geometry, recent advancements constraining the Gaussians' spread and shapes, such as 2D Gaussian Splatting, have significantly improved geometric fidelity. In this paper, we present Pixel-Aligned 1DoF Gaussian Splatting (PAGaS) that adapts the GS representation from novel view synthesis to the multi-view stereo depth task. Our key contribution is modeling a pixel's depth using one-degree-of-freedom (1DoF) Gaussians that remain tightly constrained during optimization. Unlike existing approaches, our Gaussians' positions and sizes are restricted by the back-projected pixel volumes, leaving depth as the sole degree of freedom to optimize. PAGaS produces highly detailed depths, as illustrated in Figure 1. We quantitatively validate these improvements on top of reference geometric and learning-based multi-view stereo baselines on challenging 3D reconstruction benchmarks. Code: davidrecasens.github.io/pagas

Problem

• Gaussian Splatting
• depth refinement
• multi-view stereo
• 3D reconstruction
• novel view synthesis

Innovation

• Gaussian Splatting
• 1DoF
• Depth Refinement
• Multi-view Stereo
• Pixel-Aligned