List of Datasets

Marine surface

PoLaRIS

Maritime object detection and tracking

Annotation extension of Pohang Canal with stereo RGB and thermal image labels, dynamic-object tracking IDs, and labeled LiDAR and Radar points across five day/night sequences.

Metadata: Complete Metadata needs review

Primary places to inspect, download, read about, or reproduce the dataset.

Download

Not reported

Code

Not reported

Supplementary

Not reported

Quick facts

Vehicle type
Marine surface
Environment
Outdoor
Data origin
Real-world
Sensor count
4
Ground truth
Available
Calibration
Not reported
Annotations
Available

Sensor overview

High-level sensor availability before the detailed sensor records below.

Cameras

3 reported

IMUs

Not reported

GNSS

Not reported

LiDAR

1 reported

Radar

Available; count not reported

Pitot tube

Not reported

Magnetometers

Not reported

Barometers

Not reported

UWB

Not reported

Ground truth

Reference scope, method, rate, coverage, and provenance.

Scope Method Reference system Count Rate Coverage Reference provenance
External
Measured using equipment independent of the platform’s normal onboard sensor pipeline.
Derived
Computed primarily from the dataset’s own recorded sensors.
Hybrid
Combines independent reference equipment with onboard measurements.
2D Manual annotation Not reported Not reported Not reported Partial · Dynamic Obstacles Including Moving Ships and Buoys Reference provenance
External
Measured using equipment independent of the platform’s normal onboard sensor pipeline.
Derived
Computed primarily from the dataset’s own recorded sensors.
Hybrid
Combines independent reference equipment with onboard measurements.
Derived
Reference notes
Image-space tracking IDs for dynamic objects, including small objects from first detection. No platform-pose reference is asserted.
Other Other Not reported Not reported Not reported Partial · Objects With Labeled LiDAR or Radar Returns Reference provenance
External
Measured using equipment independent of the platform’s normal onboard sensor pipeline.
Derived
Computed primarily from the dataset’s own recorded sensors.
Hybrid
Combines independent reference equipment with onboard measurements.
Derived
Reference notes
Sparse object-depth references from point-wise labels; LiDAR image projection is manually cleaned and Radar clusters are associated with labeled LiDAR points. This is not a dense per-pixel depth product.

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

Camera intrinsics and inter-sensor extrinsics are used for annotation transfer; this paper does not specify which calibration files accompany the extension.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

Distant objects have fewer LiDAR/Radar annotations than image labels because of ranging coverage limitations.

Night RGB annotation requires image restoration; TIR label transfer needs correction for extrinsic and field-of-view errors.

Cameras

stereo-rgb

Model
Not reported
Setup
Stereo
Count
1
Effective cameras
2
Modality
Rgb
Resolution
Not reported
Rate
Not reported
Shutter
Not Reported
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
Left and right image annotations; Table I reports image size 2048 by 1080 without separately documenting every optical stream. No native sensor model or release cadence is established here.

tir

Model
Not reported
Setup
Mono
Count
1
Effective cameras
1
Modality
Thermal
Resolution
Not reported
Rate
Not reported
Shutter
Not Reported
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
TIR labels are transferred from left RGB and manually refined; 16-bit imagery is converted to 8-bit with Fieldscale for annotation visualization. This does not establish the released image encoding.

IMUs

Not reported

GNSS

Not reported

LiDAR

LiDAR 1

Model or name
LiDAR 1
Count
1
Dimensions
3
Rate
Not reported
Channels
Not reported
Range
Not reported

Radar

Radar 1

Count
Not reported
Released data
Point detections / point clouds
Frame rate
Not reported
Doppler velocity
Not reported
Elevation angle
Not reported
Configured range
Not reported
Field of view
Not reported
Notes
Labeled dynamic-object Radar points in PCD; source Radar is described in 2D BEV and clustered using DBSCAN with labeled LiDAR overlap. Do not infer elevation or Doppler outputs from the hardware or coordinate projection.

Pitot tube

Not reported

Additional sensors

Not reported

Annotations

Reported annotation availability and task support.

Availability
Available
Format
txt; pcd
Class count
Not reported
Semantic segmentation Not reported
Instance segmentation Not reported
Object detection Available
Optical flow Not reported
Depth ground truth Available

Citation

Citation key and BibTeX kept at the end of the page for reference.

Citation key
choi2024polaris

BibTeX

@article{choi2024polaris,
  title={PoLaRIS Dataset: A Maritime Object Detection and Tracking Dataset in Pohang Canal},
  author={Choi, Jiwon and Cho, Dongjin and Lee, Gihyeon and Kim, Hogyun and Yang, Geonmo and Kim, Joowan and Cho, Younggun},
  year={2024},
  journal={arXiv preprint arXiv:2412.06192},
  url={https://arxiv.org/abs/2412.06192v2}
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy