Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,76 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
## UT-MM Dataset
|
2 |
+
|
3 |
+
To access our dataset, visit [Hugging Face](https://huggingface.co/datasets/neel1302/UT-MM/tree/main). To see videos of the content, please visit our [[Project Page]](https://vita-group.github.io/MM3DGS-SLAM/).
|
4 |
+
|
5 |
+
A script to convert the UT-MM rosbag data to usable .txt files is included in
|
6 |
+
`scripts/bag2data.py` at [this repository](https://github.com/VITA-Group/MM3DGS-SLAM). Note that [ROS](https://wiki.ros.org/ROS/Installation)
|
7 |
+
is required to read the camera-to-imu transformation.
|
8 |
+
|
9 |
+
```shell
|
10 |
+
python scripts/bag2data.py --path <rosbag dir path> --scene <scene name>
|
11 |
+
```
|
12 |
+
However, the transformation is constant between our datasets, which we have provided.
|
13 |
+
```
|
14 |
+
./tf/tf.txt
|
15 |
+
```
|
16 |
+
|
17 |
+
<details>
|
18 |
+
<summary><span style="font-weight: bold;">Command Line Arguments for scripts/bag2data.py</span></summary>
|
19 |
+
|
20 |
+
#### --path
|
21 |
+
|
22 |
+
Path to the directory containing rosbags
|
23 |
+
|
24 |
+
#### --scene
|
25 |
+
|
26 |
+
Name of the scene to read
|
27 |
+
|
28 |
+
</details>
|
29 |
+
<br>
|
30 |
+
|
31 |
+
### Rosbag topics
|
32 |
+
|
33 |
+
Combined, the datasets contain 8387 images, 2796 LiDAR scans, and 27971 IMU measurements. The following is a quick overview of the topics that can be found within each bag file:
|
34 |
+
|
35 |
+
```
|
36 |
+
/vrpn_client_node/Jackal_Latest/pose : geometry_msgs/PoseStamped (Pose Coordinates, 100hz)
|
37 |
+
/realsense/color/image_raw/compressed : sensor_msgs/CompressedImage (RGB Image, 30hz)
|
38 |
+
/realsense/depth/image_rect_raw : sensor_msgs/Image (Depth Image, 30hz)
|
39 |
+
/microstrain/imu/data : sensor_msgs/Imu (Imu Measurements, 100hz)
|
40 |
+
/ouster/points : sensor_msgs/PointCloud2 (LiDAR Point Clouds, 10hz)
|
41 |
+
```
|
42 |
+
|
43 |
+
To visualize the dataset, ROS is needed. Some scripts are provided in the UT_MM_Scripts directory.
|
44 |
+
|
45 |
+
```
|
46 |
+
roscore
|
47 |
+
|
48 |
+
rosrun rviz rviz -d UT_MM_Scripts/configs/jackal.rviz
|
49 |
+
|
50 |
+
rqt --clear-config --perspective-file UT_MM_Scripts/configs/rqt_plots.perspective
|
51 |
+
|
52 |
+
rosbag play --clock *.bag --pause
|
53 |
+
|
54 |
+
python3 imu_preintegration.py
|
55 |
+
```
|
56 |
+
|
57 |
+
<br>
|
58 |
+
|
59 |
+
## Citation
|
60 |
+
|
61 |
+
If you find our paper useful or interesting, please consider giving a star ⭐ and citing the following paper 📝.
|
62 |
+
|
63 |
+
```bibtex
|
64 |
+
@misc{sun2024mm3dgs,
|
65 |
+
title={MM3DGS SLAM: Multi-modal 3D Gaussian Splatting for SLAM Using Vision, Depth, and Inertial Measurements},
|
66 |
+
author={Lisong C. Sun and Neel P. Bhatt and Jonathan C. Liu and Zhiwen Fan and Zhangyang Wang and Todd E. Humphreys and Ufuk Topcu},
|
67 |
+
year={2024},
|
68 |
+
eprint={2404.00923},
|
69 |
+
archivePrefix={arXiv},
|
70 |
+
primaryClass={cs.CV}
|
71 |
+
}
|
72 |
+
```
|
73 |
+
|
74 |
+
---
|
75 |
+
license: cc
|
76 |
+
---
|