Automotive
LiDARDustX
dust-robust 3D perception
Multi-LiDAR mining-truck benchmark for dusty unstructured roads with boxes, semantic labels and tracking IDs.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- Automotive
- Environment
- Outdoor
- Data origin
- Real-world
- Sensor count
- 6
- Ground truth
- Not reported
- Calibration
- Not reported
- Annotations
- Available
Sensor overview
High-level sensor availability before the detailed sensor records below.
Ground truth
Reference scope, method, rate, coverage, and provenance.
Not reported
Calibration and synchronization
Reported calibration level, reproducibility signals, and supporting notes.
- Level
- Not reported
- Processed parameters
- Not reported
- Raw calibration data
- Not reported
- Calibration targets
- Not reported
Calibration data
Calibration reproducibility details are not reported.
Notes
Six-sensor calibration/stitching used for annotation; released parameters/raw sessions not established.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
Six LiDARs are mounted on a mining truck. Acquisition is 10 Hz and labeled keyframes are 5 Hz; they are different quantities.
Fourteen box classes, sixteen semantic classes and reduced benchmark class sets are distinct. The 180/200 sequence wording is unresolved, so aggregate sequence count stays unknown.
The paper discusses future camera/radar additions; no such released streams are asserted. No independent platform pose reference is established.
Six-sensor stitching supports labeling, but released calibration parameters/raw sessions and achieved synchronization errors remain unverified.
Cameras
Not reported
IMUs
Not reported
GNSS
Not reported
LiDAR
LS128S2
- Model or name
- LS128S2
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
CB64S1
- Model or name
- CB64S1
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
FalconK1
- Model or name
- FalconK1
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
FalconK3
- Model or name
- FalconK3
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
RS-LiDAR-M1
- Model or name
- RS-LiDAR-M1
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
Ouster2
- Model or name
- Ouster2
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
Radar
Not reported
Additional sensors
Not reported
Annotations
Reported annotation availability and task support.
- Availability
- Available
- Format
- 14 box classes, 16 semantic classes and dynamic-object tracking IDs; detection evaluation uses only truck/car/pedestrian, segmentation merges obstacle types.
- Class count
- Not reported
Semantic segmentation
Available
Instance segmentation
Not reported
Object detection
Available
Optical flow
Not reported
Depth ground truth
Not reported
Citation
Citation key and BibTeX kept at the end of the page for reference.
@misc{wei2025lidardustx,
author = {Chenfeng Wei and Qi Wu and Si Zuo and Jiahua Xu and Boyang Zhao and Zeyu Yang and Guotao Xie and Shenhong Wang},
title = {LiDARDustX: A LiDAR Dataset for Dusty Unstructured Road Environments},
year = {2025},
url = {https://arxiv.org/abs/2505.21914v1}
}Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy