List of Datasets

Automotive, UAV

V2U4Real

Vehicle-to-UAV cooperative 3D perception

Paired real vehicle/UAV LiDAR and camera recordings with four-class 3D boxes and cross-agent object identities.

Metadata: Complete Metadata source documented Released 2026

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

Supplementary

Not reported

Quick facts

Vehicle type
Automotive, UAV
Environment
Outdoor
Data origin
Real-world
Sensor count
8
Ground truth
Available
Calibration
Parameters Only
Annotations
Available

Sensor overview

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

Cameras

4 reported

IMUs

Not reported

GNSS

Not reported

LiDAR

4 reported

Radar

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 Manual annotation Manual cross-agent 3D box annotation Not reported Not reported Unknown 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
Object box poses and track identities from the recorded point clouds. Not an independent platform-pose survey. Per-object visibility limits temporal coverage.

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

Official layout lists per-frame cross-agent YAML metadata and the paper documents intrinsic/registration procedures. Complete raw calibration sessions and independently achieved errors are unverified.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

Paper prose gives 44 scenarios, while its table lists 42 sequences and differs from the nominal split/duration description. Sequence count and aggregate duration remain unknown.

Four cameras and four LiDARs are distributed between one ground vehicle and one UAV; they are not all mounted on either agent. Hardware maximum ranges are not measured collection extents.

Clock alignment and nearest-frame matching within 20 ms do not imply identical exposure times. RTK-based registration does not establish released raw GNSS or IMU observations.

Four-class boxes and cross-agent track identities derive from the point clouds. They are object references, not independent platform-pose truth; full raw calibration sessions and achieved reference accuracy remain unknown.

Cameras

Camera 1

Model
HikRobot MV-CA023-10GC
Setup
Mono
Count
3
Effective cameras
3
Modality
Rgb
Resolution
1920 x 1080
Rate
10 Hz
Shutter
Not Reported
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
Vehicle-side group: left, center and right cameras.

Camera 2

Model
HikRobot MV-CA023-10GC
Setup
Mono
Count
1
Effective cameras
1
Modality
Rgb
Resolution
1920 x 1080
Rate
10 Hz
Shutter
Not Reported
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
UAV-side camera; count is not duplicated on the vehicle.

IMUs

Not reported

GNSS

Not reported

LiDAR

Ouster-1-128

Model or name
Ouster-1-128
Count
1
Dimensions
3
Rate
10 Hz
Channels
128
Range
200 m

RoboSense Ruby-Plus-128

Model or name
RoboSense Ruby-Plus-128
Count
1
Dimensions
3
Rate
10 Hz
Channels
128
Range
250 m

RoboSense M1 Plus

Model or name
RoboSense M1 Plus
Count
1
Dimensions
3
Rate
10 Hz
Channels
Not reported
Range
200 m

Ouster-1-128

Model or name
Ouster-1-128
Count
1
Dimensions
3
Rate
10 Hz
Channels
128
Range
200 m

Radar

Not reported

Additional sensors

Not reported

Annotations

Reported annotation availability and task support.

Availability
Available
Format
3D boxes: center, dimensions and roll/pitch/yaw; cross-agent track IDs
Class count
4
Semantic segmentation Not reported
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.

Citation key
li2026v2u4real

BibTeX

@misc{li2026v2u4real,
  author = {Weijia Li and Haoen Xiang and Tianxu Wang and Shuaibing Wu and Qiming Xia and Cheng Wang and Chenglu Wen},
  title = {V2U4Real: A Real-world Large-scale Dataset for Vehicle-to-UAV Cooperative Perception},
  year = {2026},
  url = {https://arxiv.org/abs/2603.25275v1}
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy