Handheld
RoNIN
Learned pedestrian inertial navigation and body heading
Smartphone inertial sequences paired with body-mounted Tango trajectories and calibration for natural phone handling and pedestrian navigation.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- Handheld
- Environment
- Indoor
- Data origin
- Real-world
- Sensor count
- 1
- Ground truth
- Available
- Calibration
- Parameters Only
- 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
|
|---|---|---|---|---|---|---|
| 3D position + heading | Visual SLAM | Asus Zenfone AR / Tango Area Learning | Not reported | 200 Hz | 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
Per-sequence inertial bias/scale and cross-phone spatial/time alignment.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
The maintained repository says approximately half of the paper collection was not published. Paper totals of 276 sequences, 42.7 hours and 100 subjects do not describe the verified public download inventory.
The reference tracks a harness-mounted body device, not the freely handled capture phone. Visual tracking and manual relocalization correction do not establish independent absolute-position truth.
Alternative phone IMUs are not simultaneous devices. The 200 Hz synchronized sample grid does not establish separate native accelerometer/gyroscope rates.
Training-only filtered orientation can reuse reference trajectories; test-time Android orientation has different dependencies. Return-to-reference drift is not a uniform absolute-accuracy bound.
The official loader establishes inertial, Tango-pose and calibration fields. Public pressure/magnetometer arrays and raw tracking imagery were not established; they are omitted from sensor inventory.
Project/DOI requests timed out during review. The author-linked archive remains listed as the download destination; archive contents, exact public totals and achieved alignment error remain unverified.
Cameras
IMUs
Capture-phone IMU
- Model
- Not reported
- Count
- 1
- Accel / gyro
- Not reported
GNSS
LiDAR
Radar
Additional sensors
Annotations
Reported annotation availability and task support.
- Availability
- Not reported
- Format
- Not reported
- Class count
- Not reported
Citation
Citation key and BibTeX kept at the end of the page for reference.
@misc{yan2019ronin,
author = {Yan, Hang and Herath, Sachini and Furukawa, Yasutaka},
title = {RoNIN: Robust Neural Inertial Navigation in the Wild: Benchmark, Evaluations, and New Methods},
year = {2019},
eprint = {1905.12853},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/1905.12853v1}
}
Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy