Next Article in Journal
Welding Seam Trajectory Recognition for Automated Skip Welding Guidance of a Spatially Intermittent Welding Seam Based on Laser Vision Sensor
Previous Article in Journal
Dry Electrodes for Human Bioelectrical Signal Monitoring
Previous Article in Special Issue
Characterization of an Ultrasonic Local Positioning System for 3D Measurements

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

RNN-Aided Human Velocity Estimation from a Single IMU

1
Precise Positioning & Analytics Department, Fraunhofer Institute for Integrated Circuits (IIS), 90411 Nürnberg, Germany
2
Programming Systems Group, Friedrich-Alexander University (FAU), 91054 Erlangen-Nürnberg, Germany
3
Institute of Information Technology (Communication Electronics), Friedrich-Alexander University (FAU), 91054 Erlangen-Nürnberg, Germany
4
Department of Statistics, Ludwig-Maximilians-University (LMU), 80539 Munich, Germany
*
Author to whom correspondence should be addressed.
This paper is an extended version of our paper published in Feigl, T.; Kram, S.; Woller, P.; Siddiqui, R.H.; Philippsen, M.; Mutschler, C. A Bidirectional LSTM for Estimating Dynamic Human Velocities from a Single IMU. In Proceedings of the 2019 International Conference on Indoor Positioning and Indoor Navigation (IPIN), Pisa, Italy, 30 September–3 October 2019.
Sensors 2020, 20(13), 3656; https://doi.org/10.3390/s20133656
Received: 22 May 2020 / Revised: 20 June 2020 / Accepted: 24 June 2020 / Published: 29 June 2020
Pedestrian Dead Reckoning (PDR) uses inertial measurement units (IMUs) and combines velocity and orientation estimates to determine a position. The estimation of the velocity is still challenging, as the integration of noisy acceleration and angular speed signals over a long period of time causes large drifts. Classic approaches to estimate the velocity optimize for specific applications, sensor positions, and types of movement and require extensive parameter tuning. Our novel hybrid filter combines a convolutional neural network (CNN) and a bidirectional recurrent neural network (BLSTM) (that extract spatial features from the sensor signals and track their temporal relationships) with a linear Kalman filter (LKF) that improves the velocity estimates. Our experiments show the robustness against different movement states and changes in orientation, even in highly dynamic situations. We compare the new architecture with conventional, machine, and deep learning methods and show that from a single non-calibrated IMU, our novel architecture outperforms the state-of-the-art in terms of velocity (≤0.16 m/s) and traveled distance (≤3 m/km). It also generalizes well to different and varying movement speeds and provides accurate and precise velocity estimates.
Keywords: inertial navigation; motion tracking; velocity estimation; machine learning inertial navigation; motion tracking; velocity estimation; machine learning
MDPI and ACS Style

Feigl, T.; Kram, S.; Woller, P.; Siddiqui, R.H.; Philippsen, M.; Mutschler, C. RNN-Aided Human Velocity Estimation from a Single IMU. Sensors 2020, 20, 3656.

Show more citation formats Show less citations formats
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Search more from Scilit
 
Search
Back to TopTop