List of Datasets

Handheld

Oxford Spires

SLAM; 3D reconstruction; novel view synthesis

Backpack RGB/LiDAR/inertial recordings with TLS maps, registered poses and distant-view rendering benchmarks.

Metadata: Complete Metadata source documented Released 2024

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

Quick facts

Vehicle type
Handheld
Environment
Indoor, Outdoor
Data origin
Real-world
Sensor count
5
Ground truth
Available
Calibration
Partial Raw Data
Annotations
Available

Sensor overview

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

Cameras

3 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.
6DoF LiDAR SLAM Mobile-cloud registration to TLS · VILENS ICP + Leica RTC360 Not reported Not reported Partial · Core Sequences Within Tls Map 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
Claimed approximately1–2cm; validated against same reference map, not independent trajectory survey.
Reference notes
Official documentation supplies gt-tum.txt only for core paths fully covered by the TLS survey map. ICP/VILENS registration reuses normal mobile LiDAR/inertial observations; centimetre claims are not independently surveyed trajectory-wide bounds.
Other Other TLS-rendered depth · Leica RTC360 Not reported Not reported Partial · Core Sequences Within Tls Map 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
Reference notes
External TLS geometry rendered with hybrid registered camera poses; reference map coverage bounds valid depth.

Calibration and synchronization

Reported calibration level, reproducibility signals, and supporting notes.

Level
Partial Raw Data
Processed parameters
Available
Raw calibration data
Available
Calibration targets
Not reported

Reproducible calibration data

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

Notes

Hugging Face lists primary March 2024 and later camera/rig calibration sessions plus parameter files. Retain partial_raw_data: complete raw long static IMU characterization is not established.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

Twenty-four recorded paths span Oxford landmarks indoors and outdoors. Only core sequences fully covered by the surveyed TLS map receive documented reference trajectories; other paths must not inherit full reference/depth coverage.

Raw sweep-start timing differs from processed, IMU-undistorted camera-aligned clouds. FPGA/PTP clock precision does not remove laser acquisition duration or prove exact native inertial component rates.

Reference pose uses mobile LiDAR/inertial registration to external TLS geometry. Map residuals, grid resolution and approximately 1–2 cm registration claims are different quantities, without an independent universal trajectory/depth accuracy guarantee.

Raw calibration sessions and parameter files are available, but full static IMU-session release remains unknown. Conflicting site-area descriptions are omitted rather than silently resolved.

Cameras

Camera 1

Model
Sevensense Alphasense Core customized RGB cameras
Setup
Trinocular
Count
1
Effective cameras
3
Modality
Rgb
Resolution
1440 x 1080
Rate
20 Hz
Shutter
Global
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
One three-camera fisheye unit provides front/left/right RGB views; independent exposures share midpoint timestamps. Three optical views do not imply three independent trinocular systems.

IMUs

Alphasense integrated IMU

Model
Not reported
Count
1
Accel / gyro
Not reported

GNSS

Not reported

LiDAR

Hesai QT64

Model or name
Hesai QT64
Count
1
Dimensions
3
Rate
10 Hz
Channels
64
Range
Not reported

Radar

Not reported

Additional sensors

Not reported

Annotations

Reported annotation availability and task support.

Availability
Available
Format
Geometry-derived depth within registered TLS map coverage; exact downloadable depth membership is not established for every raw sequence.
Class count
Not reported
Semantic segmentation Not reported
Instance segmentation Not reported
Object detection Not reported
Optical flow Not reported
Depth ground truth Available

Citation

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

Citation key
tao2025spires

BibTeX

@article{tao2025spires,
title={The Oxford Spires Dataset: Benchmarking Large-Scale LiDAR-Visual Localisation, Reconstruction and Radiance Field Methods},
author={Tao, Yifu and Mu{\~n}oz-Ba{\~n}{\'o}n, Miguel {\'A}ngel and Zhang, Lintong and Wang, Jiahao and Fu, Lanke Frank Tarimo and Fallon, Maurice},
journal={International Journal of Robotics Research}, 
year={2025},
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy