IMU Data Processing For Inertial Aided Navigation: A Recurrent Neural Network Based Approach

  title={IMU Data Processing For Inertial Aided Navigation: A Recurrent Neural Network Based Approach},
  author={Mingming Zhang and Mingming Zhang and Yiming Chen and Mingyang Li},
  journal={2021 IEEE International Conference on Robotics and Automation (ICRA)},
In this work, we propose a novel method for performing inertial aided navigation, by using deep neural net-works (DNNs). To date, most DNN inertial navigation methods focus on the task of inertial odometry, by taking gyroscope and accelerometer readings as input and regressing for integrated IMU poses (i.e., position and orientation). While this design has been successfully applied on a number of applications, it is not of theoretical performance guarantee unless patterned motion is involved… 

Figures and Tables from this paper

Deep IMU Bias Inference for Robust Visual-Inertial Odometry With Factor Graphs

This work proposes to train a neural network to learn the true bias evolution of IMU bias explicitly, and shows that the proposed method improves state estimation in visually challenging situations across a wide range of motions by quadrupedal robots, walking humans, and drones.

Learned Inertial Odometry for Autonomous Drone Racing

This work proposes a learning- based odometry algorithm that uses an inertial measurement unit (IMU) as the only sensor modality for autonomous drone racing tasks and shows that the system is comparable to a visual-inertial odometry solution that uses a camera and exploits the known gate location and appearance.

Learning Inertial Odometry for Dynamic Legged Robot State Estimation

A novel proprioceptive state estimator for legged robots based on a learned displacement measurement from IMU data to improve state estimation in challenging scenarios where leg odometry is unreliable, such as slipping and compressible terrains is introduced.

INS Error Estimation Based on an ANFIS and Its Application in Complex and Covert Surroundings

This study proposed an inertial navigation error estimation based on an adaptive neuro fuzzy inference system (ANFIS) which can quickly and accurately output the position error of a vehicle end-to-end, which suggests that the ANFIS can considerably improve the positioning accuracy of inertial Navigation.

IMUNet: Efficient Regression Architecture for IMU Navigation and Positioning

A new architecture called IMUNet is introduced which is accurate and efficient for position estimation on edge device implementation receiving a sequence of raw IMU measurements.

PyPose: A Library for Robot Learning with Physics-based Optimization

The design goal for PyPose is to make it user-friendly,cient, and interpretable with a tidy and well-organized architecture, and it can be easily integrated into real-world robotic applications.

Prediction of Model Rocket Trajectories Using Low Cost Sensor Packages

The ability to make accurate predictions of rocket flight paths is critical for development of commercial spacecraft and other flight worthy rocket-assisted vehicles. With the advent of low cost



Denoising IMU Gyroscopes With Deep Learning for Open-Loop Attitude Estimation

This article proposes a learning method for denoising gyroscopes of Inertial Measurement Units (IMUs) using ground truth data, and estimating in real time the orientation (attitude) of a robot in

TLIO: Tight Learned Inertial Odometry

It is shown that the network, trained with pedestrian data from a headset, can produce statistically consistent measurement and uncertainty to be used as the update step in the filter, and the tightly-coupled system outperforms velocity integration approaches in position estimates, and AHRS attitude filter in orientation estimates.

LSTM-Based Zero-Velocity Detection for Robust Inertial Navigation

This work presents a method to improve the accuracy of a zero-velocity-aided inertial navigation system (INS) by replacing the standard zero- Velocity detector with a long short-term memory (LSTM) neural network, and demonstrates how this LSTM-based zero-VELocity detector operates effectively during crawling and ladder climbing.

RoNIN: Robust Neural Inertial Navigation in the Wild: Benchmark, Evaluations, & New Methods

A new benchmark containing more than 40 hours of IMU sensor data from 100 human subjects with ground-truth 3D trajectories under natural human motions, and novel neural inertial navigation architectures, making significant improvements for challenging motion cases are presented.

End-to-End Learning Framework for IMU-Based 6-DOF Odometry

This paper presents an end-to-end learning framework for performing 6-DOF odometry by using only inertial data obtained from a low-cost IMU using neural networks based on convolutional layers combined with a two-layer stacked bidirectional LSTM and a multi-task learning framework to automatically balance the weights of multiple metrics.

IONet: Learning to Cure the Curse of Drift in Inertial Odometry

It is demonstrated that IONet can generalize to estimate odometry for non-periodic motion, an extremely challenging task for existing techniques, and how deep recurrent neural networks can yield highly accurate trajectories, outperforming state-of-the-art shallow techniques on a wide range of tests and attachments.

VIMO: Simultaneous Visual Inertial Model-Based Odometry and Force Estimation

This work proposes a relative motion constraint combining the robot's dynamics and the external force in a preintegrated residual, resulting in a tightly coupled, sliding-window estimator exploiting all correlations among all variables.

On-Manifold Preintegration for Real-Time Visual--Inertial Odometry

The preintegrated inertial measurement unit model can be seamlessly integrated into a visual--inertial pipeline under the unifying framework of factor graphs and the application of incremental-smoothing algorithms and the use of a structureless model for visual measurements, which avoids optimizing over the 3-D points, further accelerating the computation.

High-precision, consistent EKF-based visual-inertial odometry

A novel, real-time EKF-based VIO algorithm is proposed, which achieves consistent estimation by ensuring the correct observability properties of its linearized system model, and performing online estimation of the camera-to-inertial measurement unit (IMU) calibration parameters.

Online IMU Intrinsic Calibration: Is It Necessary?

Observability analysis for visual-inertial navigation systems with four different inertial model variants containing intrinsic parameters that encompass one commonly used IMU model for low-cost inertial sensors theoretically confirms that the IMU intrinsics are observable given fully-excited 6-axis motion.