WildGS-SLAM: Monocular Gaussian Splatting SLAM in Dynamic Environments
Paper • 2504.03886 • Published • 10
Datasets:
Search is not available for this dataset
image image |
|---|
End of preview. Expand in Data Studio
This repository contains data for WildGS-SLAM: Monocular Gaussian Splatting SLAM in Dynamic Environments.
Paper | Project Page | Code
WildGS-SLAM accurately tracks the camera trajectory and reconstructs a 3D Gaussian map for static elements from a monocular video sequence, effectively removing dynamic components.
Datasets Used
WildGS-SLAM uses data from the following datasets:
- Wild-SLAM Mocap Dataset: (Hugging Face) Download instructions are available in the github repository.
- Wild-SLAM iPhone Dataset: (Hugging Face) Download instructions are available in the github repository.
- Bonn Dynamic Dataset: (Website) Download instructions are available in the github repository.
- TUM RGB-D (dynamic) Dataset: Download instructions are available in the github repository.
- Downloads last month
- 277