Papers by Bert-Jan van Beijnum
Journal of Neuroengineering and Rehabilitation, May 19, 2016
IEEE Transactions on Instrumentation and Measurement, 2021
Frontiers in Bioengineering and Biotechnology, Apr 12, 2018
Frontiers in Bioengineering and Biotechnology, Jan 13, 2016
IEEE Transactions on Instrumentation and Measurement, 2021
2019 International Conference on Multimodal Interaction, 2019
Adjunct of the 2019 International Conference on Multimodal Interaction, 2019
Proceedings of the International Conference on Health Informatics, 2014
Quick and easy access to performance data during matches and training sessions is important for b... more Quick and easy access to performance data during matches and training sessions is important for both players and coaches. While there are many video tagging systems available, these systems require manual efforts. In this project, we use Inertial Measurement Units (IMU) sensors strapped on the wrists of volleyball players to capture motion data and use Machine Learning techniques to model their actions and non-actions events during matches and training sessions. Analysis of the results suggests that all sensors in the IMU (i.e. magnetometer, accelerometer, barometer and gyroscope) contribute unique information in the classification of volleyball-specific actions. We demonstrate that while the accelerometer feature set provides the best Unweighted Average Recall (UAR) overall, decision fusion of the accelerometer with the magnetometer improves UAR slightly from 85.86\% to 86.9\%. Interestingly, it is also demonstrated that the non-dominant hand provides better UAR than the dominant h...
Instrumented Force Shoes™ (Xsens) was used in the INTERACTION project, to monitor gait and balanc... more Instrumented Force Shoes™ (Xsens) was used in the INTERACTION project, to monitor gait and balance measures in stroke subjects [1]. The issue of drift in foot position estimation by Inertial Measurement Units (IMUs) is corrected by Ultrasound (US), which offers relative feet distance estimation [2]. However, the US system suffers from some limitations such as synchronization between transmitter and receiver module placed on either foot, and ambience temperature. Other sensor systems including Stereo-photogrammetry, LIDAR, and magnets have been studied for relative feet distance estimation [3], [4]. However they suffer from limitations such as portability, and need for reference systems. Reflective methods using Infrared (IR) systems for distance sensing does not suffer from the above limitations and is also portable. However, studies using them either incorporate heavy systems or show large mean errors [5]. Therefore, a better distance estimation model for IR systems is required. In...
Clinical therapy following stroke aims at tackling induced impairment in motor ability, gait, and... more Clinical therapy following stroke aims at tackling induced impairment in motor ability, gait, and balance. Once transferred home, remote monitoring of subject’s performance is necessary for objective evaluation, improving mobility and preventing maladaptation. This requires a wearable and unobtrusive system capable of estimating ambulatory gait and dynamic balance measures, such as Extrapolated Centre of Mass (XCoM) and Dynamic Stability Margin (DSM). Currently, ForceShoes™ (Xsens Technologies B.V., The Netherlands) had been developed for this purpose. However, it is bulky and conspicuous. As a lightweight and inconspicuous alternative, pressure insoles (medilogic® insoles, T&T medilogic Medizintechnik GmbH, Germany) coupled with IMUs, are investigated for objective quantification of gait and dynamic balance measures. Although, to obtain such measures, 3D forces and moments are required. Linear regression models were used to model 3D forces/moments from the 1D plantar pressures meas...
2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2019
IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2020
IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2020
Sensors, 2019
Full-body motion capture typically requires sensors/markers to be placed on each rigid body segme... more Full-body motion capture typically requires sensors/markers to be placed on each rigid body segment, which results in long setup times and is obtrusive. The number of sensors/markers can be reduced using deep learning or offline methods. However, this requires large training datasets and/or sufficient computational resources. Therefore, we investigate the following research question: “What is the performance of a shallow approach, compared to a deep learning one, for estimating time coherent full-body poses using only five inertial sensors?”. We propose to incorporate past/future inertial sensor information into a stacked input vector, which is fed to a shallow neural network for estimating full-body poses. Shallow and deep learning approaches are compared using the same input vector configurations. Additionally, the inclusion of acceleration input is evaluated. The results show that a shallow learning approach can estimate full-body poses with a similar accuracy (~6 cm) to that of ...
Proceedings of the AAAI Conference on Artificial Intelligence, 2019
Previous research has shown that estimating full-body poses from a minimal sensor set using a tra... more Previous research has shown that estimating full-body poses from a minimal sensor set using a trained ANN without explicitly enforcing time coherence has resulted in output pose sequences that occasionally show undesired jitter. To mitigate such effect, we propose to improve the ANN output by combining it with a state prediction using a Kalman Filter. Preliminary results are promising, as the jitter effects are diminished. However, the overall error does not decrease substantially.
Frontiers in Physiology, 2018
Uploads
Papers by Bert-Jan van Beijnum