Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting

About

In this paper, we introduce \textbf{GS-SLAM} that first utilizes 3D Gaussian representation in the Simultaneous Localization and Mapping (SLAM) system. It facilitates a better balance between efficiency and accuracy. Compared to recent SLAM methods employing neural implicit representations, our method utilizes a real-time differentiable splatting rendering pipeline that offers significant speedup to map optimization and RGB-D rendering. Specifically, we propose an adaptive expansion strategy that adds new or deletes noisy 3D Gaussians in order to efficiently reconstruct new observed scene geometry and improve the mapping of previously observed areas. This strategy is essential to extend 3D Gaussian representation to reconstruct the whole scene rather than synthesize a static object in existing methods. Moreover, in the pose tracking process, an effective coarse-to-fine technique is designed to select reliable 3D Gaussian representations to optimize camera pose, resulting in runtime reduction and robust estimation. Our method achieves competitive performance compared with existing state-of-the-art real-time methods on the Replica, TUM-RGBD datasets. Project page: https://gs-slam.github.io/.

Chi Yan, Delin Qu, Dan Xu, Bin Zhao, Zhigang Wang, Dong Wang, Xuelong Li• 2023

Related benchmarks

TaskDatasetResultRank
Photometric RenderingReplica (room0-2, office0-4)
PSNR41.17
80
Camera TrackingReplica
Average Tracking Error0.5
68
TrackingTUM RGB-D 44 (various sequences)
Average Error23.38
41
Visual SLAMTUM RGB-D fr1 desk
ATE RMSE (cm)3.3
24
Camera TrackingTUM RGB-D fr2 xyz
ATE RMSE1.3
23
TrackingBonn RGB-D dataset
Balloon226.8
23
Camera TrackingTUM RGB-D fr1 desk
ATE RMSE3.3
23
Camera TrackingTUM RGB-D fr3 office
ATE RMSE6.6
23
Visual SLAMTUM RGB-D fr2 xyz
Translation RMSE (m)1.3
21
RenderingReplica Average
PSNR34.27
14
Showing 10 of 24 rows

Other info

Follow for update