Handheld
TUM VI
Hand-held VIO
Indoor and outdoor hand-held visual-inertial dataset with stereo grayscale cameras and partial motion-capture ground truth.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- Handheld
- Environment
- Indoor, Outdoor
- Data origin
- Real-world
- Sensor count
- 3
- Ground truth
- Available
- Calibration
- Full Raw Calibration Data
- Annotations
- Not reported
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 | Motion capture | OptiTrack · Flex13 | 16 cameras | 120 Hz | Partial · Room Sequences, Trajectory Start, Trajectory End |
Reference provenance
External
Accuracy and notes
|
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
- apriltag_grid_and_vignette_target
Reproducible calibration data
Raw calibration data is reported, so users can reproduce or inspect the calibration process.
Notes
Raw and calibrated bags, camera intrinsics/stereo extrinsics, camera-IMU and hand-eye calibration, timing offsets, IMU scale/misalignment and noise parameters, and photometric vignette calibration are provided.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
Ground truth is complete only for room sequences; longer trajectories contain 120 Hz reference poses at their start and end segments.
Short sections can have poor exposure, and the slide sequences intentionally include a visually featureless tube with very low illumination.
Apparent stuttering during rosbag playback can result from large storage chunks and is not believed to represent recorded frame drops.
Cameras
stereo-gray
- Model
- IDS uEye UI-3241LE-M-GL
- Setup
- Stereo
- Count
- 1
- Effective cameras
- 2
- Modality
- Grayscale
- Resolution
- 1024 x 1024
- Rate
- 20 Hz
- Shutter
- Global
- Lens type
- Fisheye
- HFOV
- Not reported
- VFOV
- Not reported
IMUs
Bosch BMI160
- Model
- Bosch BMI160
- Count
- 1
- Accel / gyro
- 200 / 200 Hz
GNSS
Not reported
LiDAR
Not reported
Additional sensors
Other sensor
- Model
- TAOS TSL2561
- Name
- Ambient light sensor
- Count
- 1
- Rate
- 200 Hz
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.
@inproceedings{Schubert2018,
title = {The {{TUM VI Benchmark}} for {{Evaluating Visual-Inertial Odometry}}},
booktitle = {{{IEEE International Conference}} on {{Intelligent Robots}} and {{Systems}}},
author = {Schubert, David and Goll, Thore and Demmel, Nikolaus and Usenko, Vladyslav and Stuckler, Jorg and Cremers, Daniel},
year = {2018},
eprint = {1804.06120},
pages = {1680--1687},
doi = {10.1109/IROS.2018.8593419},
archiveprefix = {arxiv},
isbn = {9781538680940}
}Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy