FootSLAM meets Adaptive Thresholding

Wahlström, Johan; Markham, Andrew; Trigoni, Niki

doi:10.1109/jsen.2020.2987813

Cited by 11 publications

(9 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…FootSLAM could solve this problem by calculating them based on the data collected at the beginning when the foot is kept still. For example, Reference [33] used an adaptive threshold in the ZUPT. However, it is impossible to leverage this technique in HeadSLAM, because the head cannot be kept completely still when the participant is in the standing pose.…”

Section: Discussionmentioning

confidence: 99%

HeadSLAM: Pedestrian SLAM with Head-Mounted Sensors

Hou

Bergmann

2022

Sensors

View full text Add to dashboard Cite

Research focused on human position tracking with wearable sensors has been developing rapidly in recent years, and it has shown great potential for application within healthcare, smart homes, sports, and emergency services. Pedestrian Dead Reckoning (PDR) with Inertial Measurement Units (IMUs) is one of the most promising solutions within this domain, as it does not rely on any additional infrastructure, whilst also being suitable for use in a diverse set of scenarios. However, PDR is only accurate for a limited period of time before unbounded errors, due to drift, affect the position estimate. Error correction can be difficult as there is often a lack of efficient methods for calibration. HeadSLAM, a method specifically designed for head-mounted IMUs, is proposed to improve the accuracy during longer tracking times (10 min). Research participants (n = 7) were asked to walk in both indoor and outdoor environments wearing head-mounted sensors, and the obtained HeadSLAM accuracy was subsequently compared to that of the PDR method. A significant difference (p < 0.001) in the average root-mean-squared error and absolute error was found between the two methods. HeadSLAM had a consist lower error across all scenarios and subjects in a 20 h walking dataset. The findings of this study show how the HeadSLAM algorithm can provide a more accurate long-term location service for head-mounted, low-cost sensors. The improved performance can support inexpensive applications for infrastructureless navigation.

show abstract

Section: Discussionmentioning

confidence: 99%

HeadSLAM: Pedestrian SLAM with Head-Mounted Sensors

Hou

Bergmann

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…We will give two examples of this approach. The first example is the use of data-driven zero-velocity detectors [3]- [10], whose output is used as input to the nonlinear filter or smoother that is used to solve (1). The second example is to replace u k in (1a) with a function learned from data [18].…”

Section: Measurementsmentioning

confidence: 99%

Three Symmetries for Data-Driven Pedestrian Inertial Navigation

Wahlström

Kok

2022

IEEE Sensors J.

Self Cite

View full text Add to dashboard Cite

The last years have seen a growing body of literature on data-driven pedestrian inertial navigation. However, despite this, it is still unclear how to efficiently combine classical models and other a priori information with existing machine learning frameworks. In this paper, we first categorize existing approaches to data-driven pedestrian inertial navigation, including approaches where a machine learning algorithm is embedded into an overarching classical framework and purely data-driven frameworks. We then propose an estimation framework where navigation estimates obtained by classical means are fed to a machine learning algorithm which is trained to correct and improve the estimates. Further, we describe three symmetries that can be used to constrain the proposed estimation framework and thereby improve its performance. These are 1) the rotational symmetry of pedestrian dynamics, 2) the rotational symmetry of the sensors, and 3) the temporal symmetry of pedestrian dynamics. To demonstrate the usefulness of the proposed framework, we use data from foot-mounted inertial sensors utilizing zero-velocity updates under mixed walking and running. Machine learning corrections are implemented using both neural networks and Gaussian processes.

show abstract

“…The less extreme dynamics at the heel strike reduce the negative effect of sensor limitations in bandwidth, sampling rate, and dynamic range, and the modeling errors are smaller since the true velocity when applying ZUPTs is closer to zero. Thus, in comparison to fast gait, modest gait speeds generally result in both better navigation performance and lower sensitivity to parameter settings [35], [59], [60]. Likewise, the performance generally improves with higher sensor quality [1], higher sampling rates [61], sensor placements on the forefoot or close to the heel [45], [57], and hard walking surfaces [57].…”

Section: Why Is There No One Thresholdmentioning

confidence: 99%

“…Adaptive thresholding Section V-A Using gait cycle segmentation Section V-B Data-driven classifiers Section V-C Other Section V-D [52], [53], [65], [72]- [81] Bayesian detectors [59], [82] Estimating speed/gait frequency [83]- [87] Using motion classification [88], [89] Other [60], [90] FSMs [91]- [93] HMMs [37], [64], [66], [94], [95] Other [43], [96] Using motion classification [69]- [71] Other [67], [68] Fig. 7.…”

Section: Robust Zero-velocity Detectorsmentioning

confidence: 99%

“…These position estimates can then be used as pseudo ground truth in the threshold calibration. In this way, it is possible to design an adaptive zero-velocity detector that does not require map information, measurements from supplementary sensors, or user input [60]. However, it should be noted that the method requires a period of time when the FootSLAM algorithm converges 9 .…”

Section: A Detectors Using Adaptive Thresholdingmentioning

confidence: 99%

See 1 more Smart Citation

Fifteen Years of Progress at Zero Velocity: A Review

Wahlström

Skog

2021

IEEE Sensors J.

Self Cite

View full text Add to dashboard Cite

Fifteen years have passed since the publication of Foxlin's seminal paper "Pedestrian tracking with shoe-mounted inertial sensors". In addition to popularizing the zero-velocity update, Foxlin also hinted that the optimal parameter tuning of the zero-velocity detector is dependent on, for example, the user's gait speed. As demonstrated by the recent influx of related studies, the question of how to properly design a robust zerovelocity detector is still an open research question. In this review, we first recount the history of foot-mounted inertial navigation and characterize the main sources of error, thereby motivating the need for a robust solution. Following this, we systematically analyze current approaches to robust zero-velocity detection, while categorizing public code and data. The article concludes with a discussion on commercialization along with guidance for future research.

show abstract

FootSLAM meets Adaptive Thresholding

Cited by 11 publications

References 33 publications

HeadSLAM: Pedestrian SLAM with Head-Mounted Sensors

HeadSLAM: Pedestrian SLAM with Head-Mounted Sensors

Three Symmetries for Data-Driven Pedestrian Inertial Navigation

Fifteen Years of Progress at Zero Velocity: A Review

Contact Info

Product

Resources

About