Automotive
MAN TruckScenes
multimodal perception; autonomous trucking
German semi-trailer truck scenes with six LiDARs, six elevation-resolving radars and tracked long-range boxes.
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
- 20
- Ground truth
- Not reported
- 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.
Not reported
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
Released intrinsic/extrinsic parameters follow photogrammetry, plane targets, angular corrections and projection checks. Cabin movement relative to the chassis limits a fixed rigid-extrinsic assumption; raw sessions are not established.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
Truck height, long-range objects, trailer occlusion and cabin flex create different geometry from passenger-car collections. Radar scan timing and rolling shutter leave measurement offsets despite clock synchronization.
Camera dimensions remain unresolved. Published ego poses fix z to zero; integrated navigation and processed cabin/chassis state are not independent 6-DoF truth or evidence of native sensor rates.
Twenty-seven annotation classes differ from the twelve evaluated detection classes. Object tracks are within scenes; approximately twenty-second clips do not establish an exact unique total recording duration.
Cameras
Camera 1
- Model
- Sekonix SF3324
- Setup
- Mono
- Count
- 4
- Effective cameras
- 4
- Modality
- Rgb
- Resolution
- Not reported
- Rate
- 10 Hz
- Shutter
- Rolling
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Four camera units provide rectified/undistorted RGB images. The paper gives 1928x1208 native versus 1980x943 processed dimensions; exact released dimensions remain unverified.
IMUs
Chassis and cabin motion units
- Model
- Xsens MTi-680G-SK
- Count
- 2
- Accel / gyro
- Not reported
GNSS
GeneSys ADMA-G-PRO+
- Model or name
- GeneSys ADMA-G-PRO+
- Count
- 1
- Rate
- Not reported
- GNSS type
- RTK
- Position output
- Available
- Velocity output
- Not reported
- Heading output
- Available
LiDAR
Hesai Pandar64
- Model or name
- Hesai Pandar64
- Count
- 2
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- 64
- Range
- Not reported
Ouster OS0
- Model or name
- Ouster OS0
- Count
- 4
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- 64
- Range
- Not reported
Radar
Continental ARS 548 RDI
- Count
- 6
- Released data
- Point detections / point clouds
- Frame rate
- 19.6 Hz
- Doppler velocity
- Available
- Elevation angle
- Available
- Configured range
- Not reported
- Field of view
- 100° H × 28° V (per radar)
- Carrier frequency
- 76 GHz
- File format
- Binary PCD
- Notes
- Six 3D radar streams with radial velocity/RCS. Nominal 20 Hz averages 19.6 Hz with drops; trailer occlusion limits coverage.
Additional sensors
Odometry
- Name
- Published ego pose
- Count
- 1
- Notes
- Processed UTM-WGS84 U32 planar translation (z always zero) plus orientation quaternion. This is navigation metadata, not independently validated 6-DoF ground truth.
Annotations
Reported annotation availability and task support.
- Availability
- Available
- Format
- nuScenes-style 3D boxes at 2 Hz, persistent within-scene track IDs, 27 classes, 15 attributes and 34 scene tags; evaluation uses a separate 12-class detection subset.
- Class count
- 27
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.
@inproceedings{fent2024truckscenes,
title = {MAN TruckScenes: A multimodal dataset for autonomous trucking in diverse conditions},
author = {Fent, Felix and Kuttenreich, Fabian and Ruch, Florian and Rizwin, Farija and Juergens, Stefan and Lechermann, Lorenz and Nissler, Christian and Perl, Andrea and Voll, Ulrich and Yan, Min and Lienkamp, Markus},
booktitle = {Advances in Neural Information Processing Systems},
editor = {A. Globerson and L. Mackey and D. Belgrave and A. Fan and U. Paquet and J. Tomczak and C. Zhang},
pages = {62062--62082},
publisher = {Curran Associates, Inc.},
url = {https://proceedings.neurips.cc/paper_files/paper/2024/file/71ac06f0f8450e7d49063c7bfb3257c2-Paper-Datasets_and_Benchmarks_Track.pdf},
volume = {37},
year = {2024}
}Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy