List of Datasets

Handheld, UGV, Automotive

SNAIL Radar

4D radar localization and mapping

44 repeated handheld, e-bike and SUV sequences with 4D radar, lidar, stereo and inertial data in campus, highway, tunnel, rain and night conditions; TLS-based and full CPGO references have distinct coverage and accuracy.

Metadata: Complete Metadata needs review

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

Download

Not reported

Supplementary

Not reported

Quick facts

Vehicle type
Handheld, UGV, Automotive
Environment
Outdoor
Data origin
Real-world
Sensor count
10
Ground truth
Available
Calibration
Parameters Only
Annotations
Not reported

Sensor overview

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

Cameras

2 reported

IMUs

3 reported

GNSS

1 reported

LiDAR

1 reported

Radar

2 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.
6DoF Other TLS reference map · Leica RTC360 Not reported 10 Hz Partial · TLS-Covered Routes; Only Beginning/End Segments on Long SUV Sequences 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.
Hybrid
Accuracy notes
Table 6 forward/backward differences measure internal consistency, not independent absolute trajectory error. Expected TLS pairwise registration within 5 cm is a different quantity.
Reference notes
Offline FAST-LIO2 localization of released lidar/IMU in a fixed map of 93 external TLS scans; forward/backward results averaged. XT32 poses in TLS frame. Scan clouds and refined TLS poses are supplied. Reference apparatus is excluded from ordinary lidar inventory.
6DoF Other Not reported Not reported Not reported Full 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.
Hybrid
Accuracy notes
Inter-session August 1 Road check reports horizontal accuracy better than 0.15 m and vertical within 1 m; not a uniform independent accuracy guarantee.
Reference notes
Full CPGO XT32 trajectories in UTM50N fuse KISS-ICP, TLS localization, GNSS/INS positions and X36D inertial factors. Intended for place recognition rather than odometry benchmarking.

Calibration and synchronization

Reported calibration level, reproducibility signals, and supporting notes.

Level
Parameters Only
Processed parameters
Available
Raw calibration data
Not reported
Calibration targets
Not reported

Parameters only

Processed calibration parameters are reported, but raw calibration data is not reported.

Notes

MATLAB scripts and ROS1 static-transform launch files provide calibration. Body frame has XT32 origin and forward/left/up axes. Manual/CAD initialization is refined with lidar-IMU calibration and correlation; November 5 ARS548 orientation is per sequence.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

TLS-based accurate trajectories cover only beginning/end segments on long SUV sequences; use full CPGO primarily for place recognition.

ZED2i image resolution changes on 2023-12-08 and some image/IMU streams have lower than nominal rates.

ARS548 and MTi3DK are absent in early sequences.

GNSS/INS heights have metre-scale spikes; CPGO vertical accuracy is only within one metre.

Enhanced Eagle point accumulation may introduce motion distortion.

Cameras

zed2i-stereo

Model
ZED2i
Setup
Stereo
Count
1
Effective cameras
2
Modality
Grayscale
Resolution
Not reported
Rate
Not reported
Shutter
Rolling
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
Nominal 20 Hz grayscale stereo, initially 1280x720; 640x360 since 2023-12-08. Some image/IMU streams fall below nominal rates. Images are JPEG; timing estimation uses 14.4 ms rolling-shutter readout.

IMUs

ZED2i-integrated IMU

Model
Not reported
Count
1
Accel / gyro
Not reported

XSens MTi3DK

Model
MTi3DK
Count
1
Accel / gyro
Not reported

Bynav X36D-integrated IMU

Model
Not reported
Count
1
Accel / gyro
Not reported

GNSS

Bynav X36D

Model or name
Bynav X36D
Count
1
Rate
100 Hz
GNSS type
RTK
Position output
Available
Velocity output
Not reported
Heading output
Not reported

LiDAR

Hesai Pandar XT32

Model or name
Hesai Pandar XT32
Count
1
Dimensions
3
Rate
10 Hz
Channels
Not reported
Range
Not reported

Radar

Continental ARS548

Count
1
Released data
Point detections / point clouds
Frame rate
15 Hz
Doppler velocity
Available
Elevation angle
Available
Configured range
Not reported
Field of view
Not reported
File format
PCD
Notes
Nominal 15 Hz. Absent before November 2023; first listed available recordings are 2023-11-05. Table 4 characteristics: max range 300 m, HFOV 120 deg, VFOV 40 deg; these are not verified used settings. Sparse clouds, approximately 150 points/frame. Per-sequence orientation on November 5 accounts for tilt.

Oculii Eagle

Count
1
Released data
Point detections / point clouds
Frame rate
15 Hz
Doppler velocity
Available
Elevation angle
Available
Configured range
Not reported
Field of view
Not reported
File format
PCD
Notes
Nominal 15 Hz; normal and enhanced accumulated point clouds plus radar_trk PointCloud topic are released. Track semantics remain unverified. Table 4 device characteristics: max range 400 m, HFOV 113 deg, VFOV 45 deg, about 7500 points/frame; do not apply enhanced density to sparse single-frame data. Accumulation can distort moving-platform clouds.

Pitot tube

Not reported

Additional sensors

Odometry

Name
ZED2i odometry
Count
1
Notes
Released /zed2i/zed_node/odom Odometry messages in TXT; estimator output rather than an additional physical sensor.

Annotations

Reported annotation availability and task support.

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

Citation

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

Citation key
Not reported

BibTeX

Not reported

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy