GaussianPro：具有渐进传播的3D高斯飞溅

摘要

最近，3D 高斯飞溅（3DGS）的出现在神经渲染领域引发了一场革命，实现了高质量渲染的实时速度。然而，3DGS 在很大程度上依赖于由运动结构（SfM）技术生成的初始化点云。在处理不可避免包含无纹理表面的大型场景时，SfM 技术总是无法在这些表面产生足够的点，并且无法为 3DGS 提供良好的初始化。因此，3DGS 遭受到困难的优化和低质量渲染。本文受经典多视图立体（MVS）技术启发，提出了一种名为 GaussianPro 的新方法，该方法应用渐进传播策略来引导 3D 高斯的密集化。与 3DGS 中使用的简单分割和克隆策略相比，我们的方法利用场景现有重建几何的先验知识和补丁匹配技术，生成具有准确位置和方向的新高斯。在大规模和小规模场景上的实验证明了我们方法的有效性，在 Waymo 数据集上，我们的方法明显优于 3DGS，PSNR 方面提高了 1.15dB。

English

The advent of 3D Gaussian Splatting (3DGS) has recently brought about a revolution in the field of neural rendering, facilitating high-quality renderings at real-time speed. However, 3DGS heavily depends on the initialized point cloud produced by Structure-from-Motion (SfM) techniques. When tackling with large-scale scenes that unavoidably contain texture-less surfaces, the SfM techniques always fail to produce enough points in these surfaces and cannot provide good initialization for 3DGS. As a result, 3DGS suffers from difficult optimization and low-quality renderings. In this paper, inspired by classical multi-view stereo (MVS) techniques, we propose GaussianPro, a novel method that applies a progressive propagation strategy to guide the densification of the 3D Gaussians. Compared to the simple split and clone strategies used in 3DGS, our method leverages the priors of the existing reconstructed geometries of the scene and patch matching techniques to produce new Gaussians with accurate positions and orientations. Experiments on both large-scale and small-scale scenes validate the effectiveness of our method, where our method significantly surpasses 3DGS on the Waymo dataset, exhibiting an improvement of 1.15dB in terms of PSNR.

GaussianPro：具有渐进传播的3D高斯飞溅

GaussianPro: 3D Gaussian Splatting with Progressive Propagation

摘要

Support