arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2403.09413cs.CV

放宽三维高斯泼溅的精确初始化约束

Relaxing Accurate Initialization Constraint for 3D Gaussian Splatting

Korea University（高丽大学）

Jaewoo Jung, Jisang Han, Honggyu An, Jiwon Kang, Seonghoon Park, Seungryong Kim

备注：
Project Page: https://ku-cvlab.github.io/RAIN-GS
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对3DGS依赖精确SfM初始化的问题，提出RAIN-GS优化策略，通过频域分析改进训练，使随机点云也能达到甚至超越精确初始化的性能。

摘要：

三维高斯泼溅（3DGS）最近在实时新视角合成和三维重建方面展示了令人印象深刻的能力。然而，3DGS严重依赖于从运动恢复结构（SfM）方法获得的精确初始化。当初始点云质量下降时，例如存在噪声或使用随机初始化的点云时，3DGS通常会遭遇较大的性能下降。为了解决这一局限性，我们提出了一种新颖的优化策略，称为RAIN-GS（放宽三维高斯泼溅的精确初始化约束）。我们的方法基于对原始3DGS优化方案的深入分析以及在频域中对SfM初始化的分析。基于我们的分析，利用简单的修改，RAIN-GS成功地从次优的点云（例如随机初始化的点云）训练三维高斯，有效放宽了对精确初始化的需求。我们通过在多个数据集上的定量和定性比较展示了我们策略的有效性，其中使用随机点云训练的RAIN-GS达到了与使用精确SfM点云训练的3DGS相当甚至更好的性能。我们的项目页面和代码可在https://ku-cvlab.github.io/RAIN-GS找到。

英文摘要：

3D Gaussian splatting (3DGS) has recently demonstrated impressive capabilities in real-time novel view synthesis and 3D reconstruction. However, 3DGS heavily depends on the accurate initialization derived from Structure-from-Motion (SfM) methods. When the quality of the initial point cloud deteriorates, such as in the presence of noise or when using randomly initialized point cloud, 3DGS often undergoes large performance drops. To address this limitation, we propose a novel optimization strategy dubbed RAIN-GS (Relaing Accurate Initialization Constraint for 3D Gaussian Splatting). Our approach is based on an in-depth analysis of the original 3DGS optimization scheme and the analysis of the SfM initialization in the frequency domain. Leveraging simple modifications based on our analyses, RAIN-GS successfully trains 3D Gaussians from sub-optimal point cloud (e.g., randomly initialized point cloud), effectively relaxing the need for accurate initialization. We demonstrate the efficacy of our strategy through quantitative and qualitative comparisons on multiple datasets, where RAIN-GS trained with random point cloud achieves performance on-par with or even better than 3DGS trained with accurate SfM point cloud. Our project page and code can be found at https://ku-cvlab.github.io/RAIN-GS.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑