Automotive
CADC
Winter driving perception, tracking and navigation
Winter surround-camera/LiDAR drives include raw and motion-corrected data, RTK navigation and instance-linked 3D cuboids.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- Automotive
- Environment
- Outdoor
- Data origin
- Real-world
- Sensor count
- 13
- 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.
| Scope | Method | Reference system | Count | Rate | Coverage |
Reference provenance
|
|---|---|---|---|---|---|---|
| 6DoF | GNSS / INS | Novatel OEM638 / Sensonor navigation | Not reported | Not reported | Unknown |
Reference provenance
Hybrid
|
|
||||||
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
Per-day calibration archives and intrinsic/extrinsic loading are documented by the devkit; raw calibration recordings are not established.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
The raw collection has 75 drives and covers 20.33 km. The labelled subset contains 7,000 synchronized samples over 12.94 km; its approximately 3 Hz sampling differs from raw sensor acquisition rates.
RTK references reuse onboard navigation with external base-station postprocessing. Validity coverage and achieved independent position/orientation error are not quantified.
Snow severity is an image/point-cloud proxy, not a calibrated snowfall measurement. Tracking UUIDs are scoped to each drive.
Raw and labelled archives have different compressed/uncompressed sizes. The separate project site was unavailable during review; the official downloader remains documented, but bulk archives were not downloaded.
Cameras
Camera 1
- Model
- Ximea MQ013CG-E2
- Setup
- Multi Camera
- Count
- 8
- Effective cameras
- 8
- Modality
- Rgb
- Resolution
- 1280 x 1024
- Rate
- 10 Hz
- Shutter
- Global
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Three forward views form a trinocular arrangement; eight physical cameras overall.
IMUs
Sensonor STIM300 inertial unit
- Model
- Sensonor STIM300
- Count
- 1
- Accel / gyro
- 100 / 100 Hz
Xsens MTi-300-AHRS inertial unit
- Model
- Xsens MTi-300-AHRS
- Count
- 1
- Accel / gyro
- 200 / 200 Hz
Xsens MTi-30-AHRS inertial unit
- Model
- Xsens MTi-30-AHRS
- Count
- 1
- Accel / gyro
- 200 / 200 Hz
GNSS
Novatel OEM638
- Model or name
- Novatel OEM638
- Count
- 1
- Rate
- Not reported
- GNSS type
- RTK
- Position output
- Available
- Velocity output
- Available
- Heading output
- Available
LiDAR
Velodyne VLP-32C
- Model or name
- Velodyne VLP-32C
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- 32
- Range
- Not reported
Radar
Additional sensors
Annotations
Reported annotation availability and task support.
- Availability
- Available
- Format
- 3D JSON cuboids with drive-level UUID, attributes, stationarity, yaw, camera and point count.
- Class count
- 10
Citation
Citation key and BibTeX kept at the end of the page for reference.
@misc{pitropov2020cadc,
author = {Pitropov, Matthew and Garcia, Danson and Rebello, Jason and Smart, Michael and Wang, Carlos and Czarnecki, Krzysztof and Waslander, Steven},
title = {Canadian Adverse Driving Conditions Dataset},
year = {2020},
eprint = {2001.10117},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2001.10117v3}
}
Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy