SMapper Light Dataset
SMapper light is a publicly available multimodal dataset collected using the SMapper platform, an open-hardware, multi-sensor device designed for SLAM (Simultaneous Localization and Mapping) research. The dataset provides 3D LiDAR, multi-camera, and IMU measurements, enabling benchmarking of visual, LiDAR, and visualโinertial SLAM methods.
๐ Applications
SMapper light dataset can be used for:
- Benchmarking LiDAR SLAM frameworks (e.g., Fast-LIO, S-Graphs)
- Benchmarking Visual SLAM frameworks (e.g., ORB-SLAM3, vS-Graphs)
- Benchmarking Multimodal SLAM frameworks
๐ Dataset Overview
Scenario | Instance | Duration | Size on Disk | Description |
---|---|---|---|---|
Indoor | IN_SMALL_01 |
01 m 29 s | 5.7 GB | Confined single-room environment |
IN_MULTI_01 |
06 m 46 s | 13.5 GB | Multi-room linear trajectory | |
IN_MULTI_02 |
07 m 07 s | 13.0 GB | Multi-room with loop closure | |
IN_LARGE_01 |
09 m 30 s | 57.9 GB | Large-scale indoor with loop closure | |
Outdoor | OUT_CAMPUS_01 |
04 m 57 s | 36.4 GB | Urban campus linear path |
OUT_CAMPUS_02 |
05 m 15 s | 37.8 GB | Urban campus circular path | |
Total | 35 m 04 s | 164.3 GB |
Info
The available topics can be found in here