Automotive
aiMotive
Long-range 3D detection, tracking and multimodal fusion
176 calibrated driving scenes provide camera, LiDAR, radar and GNSS-INS ego-motion with 14-class tracked 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
- 8
- 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 | High-precision GNSS-INS | Not reported | Not reported | Full |
Reference provenance
Derived
|
|
||||||
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
Not reportedKnown or intentional limitations
Reported constraints and characteristics to check before using the dataset.
The paper describes 176 nominal fifteen-second scenes; their arithmetic duration is not independently measured elapsed recording time.
Four physical cameras include two side fisheyes. Documented GNSS/INS ego-motion does not establish separately released raw IMU or satellite observations.
Four baseline classes differ from the fourteen native cuboid classes. Associated 2D detections are model-generated, and timing/rolling-shutter approximations remain.
Access includes an author-documented Kaggle route and a separately managed request route; complete raw calibration and all archive contents were not downloaded.
Cameras
Camera 1
- Model
- Not reported
- Setup
- Mono
- Count
- 1
- Effective cameras
- 1
- Modality
- Rgb
- Resolution
- 1920 x 1216
- Rate
- Not reported
- Shutter
- Rolling
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Rear pinhole; released export dimensions.
Camera 2
- Model
- Not reported
- Setup
- Mono
- Count
- 1
- Effective cameras
- 1
- Modality
- Rgb
- Resolution
- 1280 x 704
- Rate
- Not reported
- Shutter
- Rolling
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Front pinhole; released export dimensions.
Camera 3
- Model
- Not reported
- Setup
- Fisheye
- Count
- 2
- Effective cameras
- 2
- Modality
- Rgb
- Resolution
- 1280 x 704
- Rate
- Not reported
- Shutter
- Rolling
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Left/right fisheye; released export dimensions.
IMUs
Not reported
GNSS
Not reported
LiDAR
LiDAR 1
- Model or name
- LiDAR 1
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- 64
- Range
- Not reported
Radar
Radar 1
- Count
- 2
- Released data
- Point detections / point clouds
- Frame rate
- Not reported
- Doppler velocity
- Available
- Elevation angle
- Available
- Configured range
- Not reported
- Field of view
- Not reported
- File format
- JSON
- Notes
- Front/rear long-range radar; range, radial velocity, azimuth/elevation and reflectivity detections in JSON.
Additional sensors
Odometry
- Name
- GNSS/INS ego-motion
- Count
- 1
- Notes
- Derived frame-associated ego-motion is documented; raw GNSS and native inertial samples are not established release products.
Annotations
Reported annotation availability and task support.
- Availability
- Available
- Format
- JSON 3D cuboids, velocity, quaternion and track IDs; associated detector-generated 2D boxes
- Class count
- 14
Semantic segmentation
Not reported
Instance segmentation
Not reported
Object detection
Available
Optical flow
Not reported
Depth ground truth
Not reported
Citation
Citation key and BibTeX kept at the end of the page for reference.
@misc{matuszka2022aimotive,
author = {Tamás Matuszka and Iván Barton and Ádám Butykai and Péter Hajas and Dávid Kiss and Domonkos Kovács and Sándor Kunsági-Máté and Péter Lengyel and Gábor Németh and Levente Pető and Dezső Ribli and Dávid Szeghy and Szabolcs Vajna and Bálint Varga},
title = {aiMotive Dataset: A Multimodal Dataset for Robust Autonomous Driving with Long-Range Perception},
year = {2022},
url = {https://arxiv.org/abs/2211.09445v3}
}Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy