README.md

---
task_categories:
- robotics
language:
- en
pretty_name: SLABIM
size_categories:
- 100B<n<1T
---
<div align = "center"><h4><img src="assets/logo.png" width="5%" height="5%" /> SLABIM: </h4></div>
<div align = "center"><h4>A SLAM-BIM Coupled Dataset in HKUST Main Building</h4></div>

> Haoming Huang, [Zhijian Qiao](https://qiaozhijian.github.io/), Zehuan Yu, Chuhao Liu, [Shaojie Shen](https://uav.hkust.edu.hk/group/), Fumin Zhang and [Huan Yin](https://huanyin94.github.io/)
>
> Submitted to 2025 IEEE International Conference on Robotics & Automation

### News
* **`17 Feb 2025`:** Download Links Updated.
* **`28 Jan 2025`:** Accepted by [ICRA 2025](https://arxiv.org/abs/2502.16856).
* **`15 Sep 2024`:** We submit our paper to [IEEE ICRA](https://2025.ieee-icra.org/).

## Abstract
<div align="center"><h4>SLABIM is an open-sourced SLAM dataset that couples with BIM (Building Information Modeling).</h4></div>

<div align = "center"><img src="assets/overview.png" width="95%" /> </div>

**Features**:
+ **Large-scale Building Information Modeling**: The BIM model of this dataset is a part of the digital twin project in HKUST, 
featuring various types of offices, classrooms, lounges, and corridors.
+ **Multi-session & Multi-sensor Data**: We collect 12 sessions across different floors and regions. These sessions encompass various indoor scenarios.
+ **Dataset Validation**: To demonstrate the practicality of SLABIM, we test three different tasks: (1) LiDAR-to-BIM registration, and (2) Robot pose tracking on BIM and (3) Semantic mapping. 

## Dataset Structure
1. ```BIM/``` contains CAD files (.dxf) and mesh files (.ply) exported from the original BIM models, organized by storey and semantic tags. Users can sample
the meshes at specific densities to obtain point clouds, offering flexibility for various robotic tasks.

2. ```calibration files``` provide intrinsic camera parameters and the extrinsic parameters to the LiDAR.

3. In ```sensor data/``` directory, each session is named
```<X>F Region<Y>```, with X=1,3,4,5 and Y=1,2,3
indicating the storey and region of collection, such
as ```3F Region1```. This directory contains the **images**
and **points** produced by **camera** and **LiDAR**.

4. ```data <x>.bag```, x=0,1,2... is the **rosbag** encoding the raw information, which can be parsed using ROS tools.

5. ```sensor data/``` also contains the maps generated by SLAM, including **submap** for the LiDAR-to-BIM registration and **optimized map** by the offline mapping system.

6. ```pose_frame_to_bim.txt```, ```pose_map_to_bim.txt``` and ```pose_submap_to_bim.txt``` contains the **ground truth poses** from LiDAR scans and maps to the BIM coordinate. These poses are finely tuned using a manually
provided initial guess and local point cloud alignment.

```
SLABIM
├── BIM
│   └── <X>F
│       ├── CAD
│       │   └── <X>F.dxf
│       └── mesh
│           ├── columns.ply
│           ├── doors.ply
│           ├── floors.ply
│           └── walls.ply
├── calibration_files
│   ├── cam_intrinsics.txt
│   └── cam_to_lidar.txt
└── sensor_data
    └── <X>F_Region<Y>
        ├── images
        │   ├── data
        │   │   └── <frame_id>.png
        │   └── timestamps.txt
        ├── map
        │   ├── data
        │   │   ├── colorized.las
        │   │   └── uncolorized.ply
        │   └── pose_map_to_bim.txt
        ├── points
        │   ├── data
        │   │   └── <frame_id>.pcd
        │   ├── pose_frame_to_bim.txt
        │   └── timestamps.txt
        ├── rosbag
        │   └── data_<x>.bag
        └── submap
            ├── data
            │   └── <submap_id>.pcd
            └── pose_submap_to_bim.txt
```


<!-- ## Multi-session SLAM Dataset
<div align="left">
<img src="assets/1F.png" width=28.6% />
<img src="assets/3Fto5F.png" width=30.6% />
<img src="assets/colormap.gif" width = 39.3% >
</div> -->

## Data Acquisition Platform
The handheld sensor suite is illustrated in the Figure 1. A more detailed summary of the characteristics can be found in the Table 1.
<div align="center">
<img src="assets/sensor_suite.png" width=38.3% />
<img src="assets/collection.gif" width = 60.6% >
</div>

## Qualitative Results on SLABIM
### Global LiDAR-to-BIM Registration
Global LiDAR-to-BIM registration aims to estimate a transformation from scratch between the LiDAR submap and the BIM coordinate system. A robot can localize itself globally by aligning the online built submap to the BIM. 

<div align = "center"><img src="assets/registration.gif" width="35%" /> </div>

### Robot Pose Tracking on BIM
Different from LiDAR-to-BIM, Pose tracking requires estimating poses given the initial state and sequential measurements. 

<div align = "center"><img src="assets/pose_tracking.gif" width="35%" /> </div>

### Semantic Mapping
We deploy [FM-Fusion](https://arxiv.org/abs/2402.04555)[1] on SLABIM. For the ground truth, we convert the HKUST BIM into semantic point cloud maps using the semantic tags in BIM. Both maps contain four semantic categories: floor, wall, door, and
column, the common elements in indoor environments
<div align = "center"><img src="assets/semantic_mapping.gif" width="35%" /> </div>

[1] C. Liu, K. Wang, J. Shi, Z. Qiao, and S. Shen, “Fm-fusion: Instance-
aware semantic mapping boosted by vision-language foundation mod-
els,” IEEE Robotics and Automation Letters, 2024
## Acknowledgements
We sincerely thank Prof. Jack C. P. Cheng for generously
providing the original HKUST BIM files.

<!-- ## Citation
If you find SLABIM is useful in your research or applications, please consider giving us a star 🌟 and citing it by the following BibTeX entry. -->
<!-- ```bibtex
@ARTICLE{qiao2024g3reg,
  author={Qiao, Zhijian and Yu, Zehuan and Jiang, Binqian and Yin, Huan and Shen, Shaojie},
  journal={IEEE Transactions on Automation Science and Engineering}, 
  title={G3Reg: Pyramid Graph-Based Global Registration Using Gaussian Ellipsoid Model}, 
  year={2024},
  volume={},
  number={},
  pages={1-17},
  keywords={Point cloud compression;Three-dimensional displays;Laser radar;Ellipsoids;Robustness;Upper bound;Uncertainty;Global registration;point cloud;LiDAR;graph theory;robust estimation},
  doi={10.1109/TASE.2024.3394519}}
```
```bibtex
@inproceedings{qiao2023pyramid,
  title={Pyramid Semantic Graph-based Global Point Cloud Registration with Low Overlap},
  author={Qiao, Zhijian and Yu, Zehuan and Yin, Huan and Shen, Shaojie},
  booktitle={2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages={11202--11209},
  year={2023},
  organization={IEEE}
}
``` -->