用自生成立体图提升3D高斯点云的深度精度，减少渲染漂浮伪影。
Self-Evolving Depth-Supervised 3D Gaussian Splatting from Rendered Stereo Pairs
- 训练时自动生成虚拟立体对，动态利用立体网络提取深度线索
- 在三个主流数据集上首次评估深度准确性，显著降低几何误差
- 适合关注3D重建质量与自监督优化的研究者
3D高斯点云（GS）在准确表达三维场景几何结构方面存在显著缺陷，导致渲染深度图时出现不准确和漂浮伪影。本文针对这一问题，全面分析了在高斯原始优化过程中融合深度先验的策略，并提出一种新方法。该方法动态利用现成的立体网络，处理由GS模型自身在训练中生成的虚拟立体对，实现场景表示的持续自我优化。在三个流行数据集上的实验结果验证了该方法的有效性，首次对这些模型的深度精度进行了评估，证明其能显著改善几何重建质量。
原文摘要 · Abstract (English)
3D Gaussian Splatting (GS) significantly struggles to accurately represent the underlying 3D scene geometry, resulting in inaccuracies and floating artifacts when rendering depth maps. In this paper, we address this limitation, undertaking a comprehensive analysis of the integration of depth priors throughout the optimization process of Gaussian primitives, and present a novel strategy for this purpose. This latter dynamically exploits depth cues from a readily available stereo network, processing virtual stereo pairs rendered by the GS model itself during training and achieving consistent self-improvement of the scene representation. Experimental results on three popular datasets, breaking ground as the first to assess depth accuracy for these models, validate our findings.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。