AI-IMU Dead-Reckoning

Brossard, Martin; Barrau, Axel; Bonnabel, Silvère

doi:10.1109/tiv.2020.2980758

Cited by 199 publications

(111 citation statements)

References 42 publications

Supporting

Mentioning

109

Contrasting

Unclassified

Order By: Relevance

“…Because the loss function is on the accuracy of the filter outputs, the noise parameters are trained to produce the best state estimate, and do not necessarily best capture the measurement error model. AI-IMU [16] uses this approach to estimate IMU noise parameters and measurement uncertainties for car applications. In this work, we also combine deep learning with a Kalman filter.…”

Section: Related Workmentioning

confidence: 99%

TLIO: Tight Learned Inertial Odometry

Liu

Caruso

Ilg

et al. 2020

IEEE Robot. Autom. Lett.

151

View full text Add to dashboard Cite

In this work we propose a tightly-coupled Extended Kalman Filter framework for IMU-only state estimation. Strapdown IMU measurements provide relative state estimates based on IMU kinematic motion model. However the integration of measurements is sensitive to sensor bias and noise, causing significant drift within seconds. Recent research by Yan et al. (RoNIN) and Chen et al. (IONet) showed the capability of using trained neural networks to obtain accurate 2D displacement estimates from segments of IMU data and obtained good position estimates from concatenating them. This paper demonstrates a network that regresses 3D displacement estimates and its uncertainty, giving us the ability to tightly fuse the relative state measurement into a stochastic cloning EKF to solve for pose, velocity and sensor biases. We show that our 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. Video materials and code can be found on our project page: cathias.github.io/TLIO

show abstract

Section: Related Workmentioning

confidence: 99%

TLIO: Tight Learned Inertial Odometry

Liu

Caruso

Ilg

et al. 2020

IEEE Robot. Autom. Lett.

151

View full text Add to dashboard Cite

show abstract

“…The effect of unpredictable nonuniform noise as well as external environmental conditions is also inevitable [35]. To enhance the accuracy of localization, the solutions found in the literature can be classified into: (1) controlling the environment under investigation [36], (2) sensor data fusion [37], [38], (3) improving measurement covariane estimation [39], [30], [35], [40], or (4) correcting measurement errors, which can be further classified into classical [41], [42], [43], [44] and learning approaches [16], [45], [34], [46], [17], [47].…”

Section: Enhancing Slam Estimation Accuracymentioning

confidence: 99%

“…More particularly, the approach targets unmodeled, unpredictable noise patterns that are experienced in marine environments. A recent noise model learning approach was proposed in [40] where a deep neural network was trained to estimate the covariance of inertial measurements, which are then used in an extended Kalman filter to perform localization. Evaluation results demonstrated the applicability of the approach, yet in its current version, it works for inertial sensors only.…”

Section: Enhancing Slam Estimation Accuracymentioning

confidence: 99%

A Stacked LSTM-Based Approach for Reducing Semantic Pose Estimation Error

Azzam

Alkendi

Taha³

et al. 2021

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

Achieving high estimation accuracy is significant for semantic simultaneous localization and mapping (SLAM) tasks. Yet, the estimation process is vulnerable to several sources of error including limitations of the instruments used to perceive the environment, shortcomings of the employed algorithm, environmental conditions, or other unpredictable noise. In this paper, a novel stacked long short term memory (LSTM) based error reduction approach is developed to enhance the accuracy of semantic SLAM in presence of such error sources. Training and testing datasets were constructed through simulated and realtime experiments. The effectiveness of the proposed approach was demonstrated by its ability to capture and reduce semantic SLAM estimation errors in training and testing datasets. Quantitative performance measurement was carried out using the absolute trajectory error (ATE) metric. The proposed approach was compared to vanilla and bidirectional LSTM networks, shallow and deep neural networks, and to support vector machines. The proposed approach outperforms all other structures and was able to significantly improve the accuracy of semantic SLAM. To further verify the applicability of the proposed approach, it was tested on real-time sequences from the TUM RGB-D dataset, where it was able to improve the estimated trajectories.

show abstract

“…The major downside of this approach is that the accuracy of these models is highly dependent on the data used to train them. Proponents of the psuedo-measurement approach argue that the best way to incorporate learned models is by adding them as additional measurements to an existing navigation system (e.g., a Kalman filter) [38,49,52,53,55,58]. The benefit of this approach is that the existing navigation system is augmented rather than replaced; however, the challenge of this approach comes from determining the details of how the learned model(s) will be integrated into the existing system.…”

Section: Related Workmentioning

confidence: 99%

Localization of Biobotic Insects Using Low-Cost Inertial Measurement Units

Cole

Bozkurt

Lobatón

2020

Sensors

View full text Add to dashboard Cite

Disaster robotics is a growing field that is concerned with the design and development of robots for disaster response and disaster recovery. These robots assist first responders by performing tasks that are impractical or impossible for humans. Unfortunately, current disaster robots usually lack the maneuverability to efficiently traverse these areas, which often necessitate extreme navigational capabilities, such as centimeter-scale clearance. Recent work has shown that it is possible to control the locomotion of insects such as the Madagascar hissing cockroach (Gromphadorhina portentosa) through bioelectrical stimulation of their neuro-mechanical system. This provides access to a novel agent that can traverse areas that are inaccessible to traditional robots. In this paper, we present a data-driven inertial navigation system that is capable of localizing cockroaches in areas where GPS is not available. We pose the navigation problem as a two-point boundary-value problem where the goal is to reconstruct a cockroach’s trajectory between the starting and ending states, which are assumed to be known. We validated our technique using nine trials that were conducted in a circular arena using a biobotic agent equipped with a thorax-mounted, low-cost inertial measurement unit. Results show that we can achieve centimeter-level accuracy. This is accomplished by estimating the cockroach’s velocity—using regression models that have been trained to estimate the speed and heading from the inertial signals themselves—and solving an optimization problem so that the boundary-value constraints are satisfied.

show abstract

AI-IMU Dead-Reckoning

Cited by 199 publications

References 42 publications

TLIO: Tight Learned Inertial Odometry

TLIO: Tight Learned Inertial Odometry

A Stacked LSTM-Based Approach for Reducing Semantic Pose Estimation Error

Localization of Biobotic Insects Using Low-Cost Inertial Measurement Units

Contact Info

Product

Resources

About