arXiv 2508.070032025-08-09cs.RO

EGS-SLAM：基于事件的 RGB-D 高斯泼溅 SLAM

EGS-SLAM: RGB-D Gaussian Splatting SLAM with Events

School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore（南洋理工大学电子与电气工程学院）；School of Instrumentation and Optoelectronics Engineering, Beihang University, Beijing, China（北京航空航天大学仪器与光电工程学院）

Siyu Chen, Shenghai Yuan, Thien-Minh Nguyen, Zhuyu Huang, Chenyang Shi, Jin Jing, Lihua Xie

备注：: Accepted by IEEE RAL

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出 EGS-SLAM 框架，通过融合事件数据与 RGB-D 输入并建模相机连续曝光轨迹，有效解决运动模糊问题，实现了高精度的稳健跟踪与高保真 3D 高斯泼溅重建。

摘要：

Gaussian Splatting SLAM (GS-SLAM) 相比传统 SLAM 方法有显著改进，实现了传统方法通常难以达到的逼真 3D 重建。然而，现有的 GS-SLAM 系统在现实场景中常见的持续且严重的运动模糊下表现不佳，导致跟踪精度显著下降和 3D 重建质量受损。为了解决这一局限性，我们提出了 EGS-SLAM，一种新颖的 GS-SLAM 框架，将事件数据与 RGB-D 输入融合，以同时减少图像中的运动模糊并弥补事件流的稀疏和离散特性，从而实现稳健的跟踪和高保真 3D Gaussian Splatting 重建。具体而言，我们的系统显式地建模了相机在曝光期间的连续轨迹，在统一的 3D Gaussian Splatting 场景上支持事件和模糊感知的跟踪与建图。此外，我们引入了一个可学习的相机响应函数来对齐事件和图像的动态范围，并引入无事件损失以抑制重建过程中的振铃伪影。我们在一个包含具有显著运动模糊的合成和真实世界序列的新数据集上验证了我们的方法。大量实验结果表明，EGS-SLAM 在轨迹精度和逼真 3D Gaussian Splatting 重建方面始终优于现有的 GS-SLAM 系统。源代码将在 https://github.com/Chensiyu00/EGS-SLAM 公开。

英文摘要：

Gaussian Splatting SLAM (GS-SLAM) offers a notable improvement over traditional SLAM methods, enabling photorealistic 3D reconstruction that conventional approaches often struggle to achieve. However, existing GS-SLAM systems perform poorly under persistent and severe motion blur commonly encountered in real-world scenarios, leading to significantly degraded tracking accuracy and compromised 3D reconstruction quality. To address this limitation, we propose EGS-SLAM, a novel GS-SLAM framework that fuses event data with RGB-D inputs to simultaneously reduce motion blur in images and compensate for the sparse and discrete nature of event streams, enabling robust tracking and high-fidelity 3D Gaussian Splatting reconstruction. Specifically, our system explicitly models the camera's continuous trajectory during exposure, supporting event- and blur-aware tracking and mapping on a unified 3D Gaussian Splatting scene. Furthermore, we introduce a learnable camera response function to align the dynamic ranges of events and images, along with a no-event loss to suppress ringing artifacts during reconstruction. We validate our approach on a new dataset comprising synthetic and real-world sequences with significant motion blur. Extensive experimental results demonstrate that EGS-SLAM consistently outperforms existing GS-SLAM systems in both trajectory accuracy and photorealistic 3D Gaussian Splatting reconstruction. The source code will be available at https://github.com/Chensiyu00/EGS-SLAM.

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。