List of Datasets

Handheld

Hilti-Oxford

Multisensor SLAM with sparse surveyed position references

Sixteen handheld runs from the 2022 Hilti challenge with LiDAR, five fisheye cameras and IMU, sparse surveyed positions, and dense references for three runs.

Metadata: Complete Metadata source documented Released 2022

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

Supplementary

Not reported

Quick facts

Vehicle type
Handheld
Environment
Indoor, Outdoor
Data origin
Real-world
Sensor count
7
Ground truth
Available
Calibration
Partial Raw Data
Annotations
Not reported

Sensor overview

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

Cameras

5 reported

IMUs

1 reported

GNSS

Not reported

LiDAR

1 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.
3D position Surveyed markers Z+F Imager5016 surveyed crosshairs Not reported Not reported Partial 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.
External
Accuracy notes
Manual placement below 1 mm is claimed; survey, placement and extrinsic uncertainties are distinct.
Reference notes
Sparse surveyed 3D positions are supplied for all 16 sequences, expressed in the IMU frame using the needle-tip transform. Only a few stationary instants per run are measured; no dense orientation reference is implied.
6DoF LiDAR SLAM VILENS localization in external survey map Not reported Not reported Partial · Exp14, Exp16, Exp18 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
A general 1–2 cm claim for the dense additional-sequence references is distinct from sparse manual-placement accuracy; no uniform independently measured trajectory bound is established.
Reference notes
Dense six-DoF trajectories for Exp14, Exp16 and Exp18 reuse released LiDAR and IMU through VILENS localization in an external survey map. Outputs are expressed in the IMU frame.

Calibration and synchronization

Reported calibration level, reproducibility signals, and supporting notes.

Level
Partial Raw Data
Processed parameters
Available
Raw calibration data
Available
Calibration targets
7×12 AprilTag board with 15 cm tags

Reproducible calibration data

Raw calibration data is reported, so users can reproduce or inspect the calibration process.

Notes

Camera–IMU Kalibr recordings and a stationary 90-minute IMU recording are supplied. No separate LiDAR–IMU extrinsic calibration was undertaken; mechanical alignment is distinct from calibrated extrinsics.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

The 2022 Hilti-Oxford release comprises eight challenge and eight additional runs at Schaan and Oxford. It is distinct from the earlier Hilti SLAM release and later challenge editions.

Sparse surveyed position controls cover all 16 runs at discrete stationary instants. Dense six-DoF references are available only for Exp14, Exp16 and Exp18 and depend on the released LiDAR and IMU.

The claimed submillimeter needle placement is not a bound on full-trajectory error. Survey, manual placement and sensor extrinsic uncertainties remain separate; a possible few-millimeter control-point error is acknowledged.

Calibration includes some raw camera–IMU and IMU-noise recordings, but no separate LiDAR–IMU extrinsic calibration. It is classified as partial raw calibration data.

Cameras

Camera 1

Model
Alphasense Core Development Kit
Setup
Stereo
Count
1
Effective cameras
2
Modality
Grayscale
Resolution
720 x 540
Rate
40 Hz
Shutter
Global
Lens type
Fisheye
HFOV
Not reported
VFOV
Not reported
Notes
Front pair within five-view rig.

Camera 2

Model
Alphasense Core Development Kit
Setup
Mono
Count
3
Effective cameras
3
Modality
Grayscale
Resolution
720 x 540
Rate
40 Hz
Shutter
Global
Lens type
Fisheye
HFOV
Not reported
VFOV
Not reported
Notes
Two lateral and one upward views.

IMUs

Bosch BMI085

Model
Bosch BMI085
Count
1
Accel / gyro
400 / 400 Hz

GNSS

Not reported

LiDAR

Hesai PandarXT-32

Model or name
Hesai PandarXT-32
Count
1
Dimensions
3
Rate
10 Hz
Channels
32
Range
120 m

Radar

Not reported

Additional sensors

Not reported

Handheld, UGV

Hilti SLAM Challenge

Original 2021 twelve-sequence camera, dual-lidar and IMU collection with sparse total-station position and lab/arena motion-capture references.

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
zhang2023hilti

BibTeX

@article{zhang2023hilti,
  author = {Zhang, Lintong and Helmberger, Michael and Fu, Lanke Frank Tarimo and Wisth, David and Camurri, Marco and Scaramuzza, Davide and Fallon, Maurice},
  title = {{Hilti-Oxford} Dataset: A Millimeter-Accurate Benchmark for Simultaneous Localization and Mapping},
  journal = {IEEE Robotics and Automation Letters},
  volume = {8},
  number = {1},
  pages = {408--415},
  year = {2023},
  doi = {10.1109/LRA.2022.3226077}
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy