| cff-version: 1.2.0 |
| message: >- |
| If you use this dataset, please cite the upstream VBR paper (the source) and |
| the TartanIMU paper (the target loader / foundation model that this dataset |
| is built to feed). Citing this re-projection itself is optional but appreciated. |
| |
| title: "VBR → TartanIMU (car + handheld IMU, retargetted)" |
| abstract: >- |
| A re-projection of the IMU + ground-truth-pose channels of the VBR (Vision |
| Benchmark in Rome) dataset into the schema expected by the TartanIMU |
| AirLabNPZSequence data loader. Includes 8 sequences (4 car + 4 handheld) |
| recorded by the RVP Group at Sapienza University of Rome, totalling |
| 1.34 M IMU samples at uniform 200 Hz across 1.87 hours of recording. |
| Each NPZ contains body-frame IMU (accel + gyro) with gravity, world-frame |
| position, and body-to-world quaternion at strictly 200 Hz with float64 |
| timestamps. Drop-in compatible with TartanIMU training; no monkey-patch |
| required because sequence path layout disambiguates motion_type via the |
| loader's substring matcher. |
| |
| type: dataset |
| version: "1.0.0" |
| date-released: "2026-05-18" |
| license: BSD-3-Clause |
| license-url: "https://github.com/rvp-group/vbr-devkit/blob/main/LICENSE" |
|
|
| keywords: |
| - inertial measurement unit |
| - IMU |
| - inertial odometry |
| - dead reckoning |
| - autonomous driving |
| - handheld |
| - SLAM |
| - foundation model |
| - VBR |
| - vision benchmark in Rome |
| - tartanimu |
| - SBG Ellipse-E |
|
|
| references: |
| - type: conference-paper |
| title: "VBR: A Vision Benchmark in Rome" |
| authors: |
| - family-names: Brizi |
| given-names: Leonardo |
| - family-names: Bartoccioni |
| given-names: Florent |
| conference: |
| name: "IEEE International Conference on Robotics and Automation (ICRA)" |
| year: 2024 |
| url: "https://rvp-group.net/slam-dataset.html" |
|
|
| - type: conference-paper |
| title: "Tartan IMU: A Light Foundation Model for Inertial Positioning in Robotics" |
| authors: |
| - family-names: Zhao |
| given-names: Shibo |
| - family-names: Zhou |
| given-names: Sifan |
| - family-names: Blanchard |
| given-names: Raphael |
| - family-names: Qiu |
| given-names: Yuheng |
| - family-names: Wang |
| given-names: Wenshan |
| - family-names: Scherer |
| given-names: Sebastian |
| conference: |
| name: "Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)" |
| year: 2025 |
| url: "https://superodometry.com/tartanimu" |
|
|
