Neural Inertial Localization

CVPR 2022

Sachini Herath 1, David Caruso 2,
Chen Liu 2, Yufan Chen 2, Yasutaka Furukawa 1

1 Simon Fraser University   2 Reality Labs, Meta

Preprint Paper Code Data

This paper proposes the inertial localization problem, the task of estimating the absolute location from a sequence of inertial sensor measurements. This is an exciting and unexplored area of indoor localization research, where we present a rich dataset with 53 hours of inertial sensor data and the associated ground truth locations. We developed a solution, dubbed neural inertial localization NILoc which 1) uses a neural inertial navigation technique to turn inertial sensor history to a sequence of velocity vectors; then 2) employs a transformer-based neural architecture to find the device location from the sequence of velocities. We only use an IMU sensor, which is energy efficient and privacy preserving compared to WiFi, cameras, and other data sources. Our approach is significantly faster and achieves competitive results even compared with state-of-the-art methods that require a floorplan and run 20 to 30 times slower. We share our code, model and data.

Dataset

NILoc inertial localization dataset contains 53 hours of motion/trajectory data from two university buildings and one office space. Each scene spans a flat floor and we share IMU data and ground-truth locations based on Visual Inertial SLAM for each trajectory.

[Download]

Dataset Environment Full Dataset statistics
Dimensions
(m^2)		 resolution
(pixels per meter)		 trajectories subjects duration (h) length(km)
University A 62.8 x 84.4 2.5 151 52 25.57 65.35
University B 57.6 x 147.2 2.5 91 3 14.64 59.63
Office C 38.4 x 11.2 10 77 1 12.8 25.65

Architecture

We use two branch transformer architecture to estimate location likelihood from IMU velocity input.

[Github Repo]

Related Projects

Check out our related projects on the topic of inertial navigation and localization!

Fusion-DHL: WiFi, IMU, and Floorplan Fusion for Dense History of Locations in Indoor Environments (ICRA 2021)
A multi-modal sensor fusion algorithm that fuses WiFi, IMU, and floorplan information to infer an accurate and dense location history in indoor environments. The algorithm uses 1) an inertial navigation algorithm to estimate a relative motion trajectory from IMU sensor data; 2) a WiFi-based localization API in industry to obtain positional constraints and geo-localize the trajectory; and 3) a convolutional neural network to refine the location history to be consistent with the floorplan.
RoNIN: Robust Neural Inertial Navigation in the Wild (ICRA 2020)
The research focus on data-driven inertial navigation, where the task is the estimation of positions and orientations of a moving subject from a sequence of IMU sensor measurements. We present 1) a new benchmark of IMU sensor data and ground-truth 3D trajectories under natural human motions; 2) neural inertial navigation architectures, making significant improvements for challenging motion cases; and 3) comprehensive evaluations of the competing methods and datasets.

Bibtex

@inproceedings{herath2020niloc, author = {Herath, Sachini and Caruso, David and Liu, Chen, and Chen, Yufan and Furukawa, Yasutaka}, title = {Neural Inertial Localization}, url = {https://arxiv.org/abs/2203.15851}, publisher = {arXiv}, year={2022}}

Acknowledgement

The research is supported by NSERC Discovery Grants, NSERC Discovery Grants Accelerator Supplements, DND/NSERC Discovery Grant Supplement, and John R. Evans Leaders Fund (JELF).