List of Datasets

Handheld, Automotive

VBR

visual odometry; SLAM

Handheld and car routes across Rome with stereo/LiDAR/inertial streams and globally refined reference trajectories.

Metadata: Complete Metadata source documented Released 2024

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

Supplementary

Not reported

Quick facts

Vehicle type
Handheld, Automotive
Environment
Indoor, Outdoor
Data origin
Real-world
Sensor count
6
Ground truth
Available
Calibration
Full Raw Calibration Data
Annotations
Not reported

Sensor overview

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

Cameras

2 reported

IMUs

1 reported

GNSS

1 reported

LiDAR

2 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 RTK + LiDAR geometric/photometric bundle adjustment Not reported Not reported Partial · Training Recordings 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
Accuracy notes
Six surveyed markers check relative geometry on a 1.5 km validation path; the reported approximately 3 cm is not an independent whole-corpus trajectory/orientation bound.
Reference notes
RTK priors and recorded LiDAR feed geometric/photometric bundle adjustment. Training references are distributed, benchmark test truth is withheld; normal inputs are reused.

Calibration and synchronization

Reported calibration level, reproducibility signals, and supporting notes.

Level
Full Raw Calibration 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

The paper explicitly releases per-sequence checkerboard, plane-matching and hand-eye calibration recordings. Maintained downloader documents vbr_calib.yaml with the bags and training reference; complete raw sessions were not independently downloaded.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

The maintained downloader enumerates sixteen recordings across six route groups: eight training and eight test records. Paper six-route sums of 40.744 km and 12,710 seconds are retained as historical scope, not assumed totals for all sixteen downloads.

Handheld and car LiDAR/camera configurations are alternatives. The camera model conflict remains unresolved, and native inertial/GNSS sampling is distinct from published output cadence.

Training references reuse LiDAR/RTK inputs and test truth is withheld. Relative surveyed-marker validation does not certify absolute position/orientation accuracy for all trajectories.

Timing depends on subsystem triggering, PPS, offline offset correction and clock drift. Current FTP archive reachability and complete calibration-file inventory were not exhaustively checked.

Cameras

Camera 1

Model
Not reported
Setup
Stereo
Count
1
Effective cameras
2
Modality
Rgb
Resolution
1388 x 700
Rate
Not reported
Shutter
Global
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
One global-shutter stereo system with Bayer images and approximately 50 cm baseline. Paper Manta G-145 versus G-125B/C model claims conflict; camera model stays unknown. Published cadence varies: 30 Hz handheld and 20 Hz car.

IMUs

SBG integrated inertial unit

Model
SBG Ellipse-E
Count
1
Accel / gyro
Not reported

GNSS

GNSS 1

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

LiDAR

Ouster OS0-128

Model or name
Ouster OS0-128
Count
1
Dimensions
3
Rate
10 Hz
Channels
128
Range
Not reported

Ouster OS1-64

Model or name
Ouster OS1-64
Count
1
Dimensions
3
Rate
20 Hz
Channels
64
Range
Not reported

Radar

Not reported

Additional sensors

Not reported

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
brizi2024vbr

BibTeX

@misc{brizi2024vbr,
  author = {{Leonardo Brizi} and {Emanuele Giacomini} and {Luca Di Giammarino} and {Simone Ferrari} and {Omar Salem} and {Lorenzo De Rebotti} and {Giorgio Grisetti}},
  title = {VBR: A Vision Benchmark in Rome},
  year = {2024},
  url = {https://arxiv.org/abs/2404.11322v1}
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy