Towards Real-Time Gaussian Splatting: Accelerating 3DGS through Photometric SLAM

TL;DR

Preliminary experiments show that using Direct Sparse Odometry point cloud outputs, as opposed to standard structure-from-motion methods, significantly shortens the training time needed to achieve high-quality renders, which enables the development of 3DGS-integrated SLAM systems that operate in real-time on mobile hardware.

Abstract

Initial applications of 3D Gaussian Splatting (3DGS) in Visual Simultaneous Localization and Mapping (VSLAM) demonstrate the generation of high-quality volumetric reconstructions from monocular video streams. However, despite these promising advancements, current 3DGS integrations have reduced tracking performance and lower operating speeds compared to traditional VSLAM. To address these issues, we propose integrating 3DGS with Direct Sparse Odometry, a monocular photometric SLAM system. We have done preliminary experiments showing that using Direct Sparse Odometry point cloud outputs, as opposed to standard structure-from-motion methods, significantly shortens the training time needed to achieve high-quality renders. Reducing 3DGS training time enables the development of 3DGS-integrated SLAM systems that operate in real-time on mobile hardware. These promising initial findings suggest further exploration is warranted in combining traditional VSLAM systems with 3DGS.

Figures (2)

• Figure 1: Renders of Replica Room 0 at different training stages. Notice the high quality of renders even at earlier iterations.
• Figure 2: Peak Signal to Noise Ratio (PSNR) of training of Room 0 of the Replica dataset over time. Higher PSNR is better. Average taken over ten runs.