Handheld, UAV

FIReStereo

thermal stereo; depth perception

Thermal stereo forest/urban recordings with smoke, darkness and LiDAR-derived depth on smokeless portions.

Metadata: Complete Metadata source documented Released 2024

Quick facts

Vehicle type
Handheld, UAV
Environment
Outdoor
Data origin
Real-world
Sensor count
4
Ground truth
Available
Calibration
Parameters Only
Annotations
Available

Sensor overview

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

Ground truth

Reference scope, method, rate, coverage, and provenance.

Ground-truth references for FIReStereo
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.
Other LiDAR SLAM Projected Faster-LIO map 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.
Derived
Reference notes
Input LiDAR/IMU map with interpolated poses and thermal projection;35706 smokeless pairs only, no smoky depth reference.

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

Heated checkerboard Kalibr; manual camera–LiDAR correction; start-only flat-field correction.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

Thermal data are 16-bit raw except Firesgl, which is logged as 8-bit RGB. Native released camera cadence remains unknown; the maximum camera specification is not used as its rate.

Projected depth is derived from input LiDAR/IMU mapping on smokeless portions. Smoky fire sequences do not have the same depth reference, and LiDAR/IMU bags are documented for Frick and Hawkins.

The repository requests a short survey and provides downloader scripts. Calibration parameters are linked, while complete raw sessions and achieved synchronization/reference errors remain unverified.

Cameras

Camera 1

Model
FLIR Boson+640
Setup
Stereo
Count
1
Effective cameras
2
Modality
Thermal
Resolution
640 x 512
Rate
Not reported
Shutter
Not Reported
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
One stereo system with two thermal heads and a 24.6 cm baseline. Most collections contain 16-bit raw images; Firesgl is logged as 8-bit RGB. Rated 30 fps does not establish the delivered native cadence.

IMUs

Epson G365

Model
Epson G365
Count
1
Accel / gyro
Not reported

GNSS

Not reported

LiDAR

Velodyne VLP16

Model or name
Velodyne VLP16
Count
1
Dimensions
3
Rate
10 Hz
Channels
16
Range
Not reported

Radar

Not reported

Additional sensors

Not reported

Annotations

Reported annotation availability and task support.

Availability
Available
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 Available

Citation

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

Citation key
dhrafani2024firestereo

BibTeX

@misc{dhrafani2024firestereo,
  author = {Devansh Dhrafani and Yifei Liu and Andrew Jong and Ukcheol Shin and Yao He and Tyler Harp and Yaoyu Hu and Jean Oh and Sebastian Scherer},
  title = {FIReStereo: Forest InfraRed Stereo Dataset for UAS Depth Perception in Visually Degraded Environments},
  year = {2024},
  url = {https://arxiv.org/abs/2409.07715v1}
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy