Stride Length Estimation Research Articles

Uncertainty-aware ensemble model for stride length estimation in gait analysis

Deep neural network (DNN) models have become highly effective tools for function estimation in regression tasks, with applications in various domains, including human gait stride length estimation. Several studies have developed DNN models that use gait cycle data from wearable devices equipped with inertial measurement units to accurately predict stride lengths within a cycle. However, many of these deep learning approaches do not quantify predictive uncertainty and fail to consider individual gait characteristics. To address these limitations, we introduced an ensemble model based on a heteroscedastic neural network that quantifies the uncertainty in stride length predictions. Additionally, we proposed a novel stride training strategy that uses the average stride instead of individual strides to optimize the training process and improve model efficiency. Our extensive performance evaluations demonstrate the robustness and adaptability of our model, accurately predicting both the uncertainty of individual stride lengths and the specific stride lengths for each gait cycle. Our study represents a significant advancement in stride length estimation, addressing the challenges of uncertainty quantification and capturing the unique aspects of human gait. The proposed model has considerable potential for practical applications in gait analysis and related fields.

Gait Stride Length Estimation Using Embedded Machine Learning.

Spatiotemporal gait parameters, e.g., gait stride length, are measurements that are classically derived from instrumented gait analysis. Today, different solutions are available for gait assessment outside the laboratory, specifically for spatiotemporal gait parameters. Such solutions are wearable devices that comprise an inertial measurement unit (IMU) sensor and a microcontroller (MCU). However, these existing wearable devices are resource-constrained. They contain a processing unit with limited processing and memory capabilities which limit the use of machine learning to estimate spatiotemporal gait parameters directly on the device. The solution for this limitation is embedded machine learning or tiny machine learning (tinyML). This study aims to create a machine-learning model for gait stride length estimation deployable on a microcontroller. Starting from a dataset consisting of 4467 gait strides from 15 healthy people, measured by IMU sensor, and using state-of-the-art machine learning frameworks and machine learning operations (MLOps) tools, a multilayer 1D convolutional float32 and int8 model for gait stride length estimation was developed. The developed float32 model demonstrated a mean accuracy and precision of 0.23 ± 4.3 cm, and the int8 model demonstrated a mean accuracy and precision of 0.07 ± 4.3 cm. The memory usage for the float32 model was 284.5 kB flash and 31.9 kB RAM. The int8 model memory usage was 91.6 kB flash and 13.6 kB RAM. Both models were able to be deployed on a Cortex-M4F 64 MHz microcontroller with 1 MB flash memory and 256 kB RAM. This study shows that estimating gait stride length directly on a microcontroller is feasible and demonstrates the potential of embedded machine learning, or tinyML, in designing wearable sensor devices for gait analysis.

Accurate Stride-Length Estimation Based on LT-StrideNet for Pedestrian Dead Reckoning Using a Shank-Mounted Sensor.

Pedestrian dead reckoning (PDR) is a self-contained positioning technology and has been a significant research topic in recent years. Pedestrian-stride-length estimation is the core part of the PDR system and directly affects the performance of the PDR. The current stride-length-estimation method is difficult to adapt to changes in pedestrian walking speed, which leads to a rapid increase in the error of the PDR. In this paper, a new deep-learning model based on long short-term memory (LSTM) and Transformer, LT-StrideNet, is proposed to estimate pedestrian-stride length. Next, a shank-mounted PDR framework is built based on the proposed stride-length-estimation method. In the PDR framework, the detection of pedestrian stride is achieved by peak detection with a dynamic threshold. An extended Kalman filter (EKF) model is adopted to fuse the gyroscope, accelerometer, and magnetometer. The experimental results show that the proposed stride-length-estimation method can effectively adapt to changes in pedestrian walking speed, and our PDR framework has excellent positioning performance.

Visual Inertial Odometry-Based Gait Analysis Using Waist-Attached RGB-D Camera and Inertial Sensors

In this article, a visual–inertial odometry algorithm is proposed to estimate walking stride length and reconstruct walking trajectory. Depth and color image data from a downward-looking waist-mounted red green blue-depth (RGB-D) camera is fused with its internal inertial measurement unit (IMU) data in an estimation algorithm to perform foot detection and position estimation. Floor plane and foot positions in stance phases are calculated and used as landmarks to construct measurement equations for updating the filter. A smoothing problem is formulated as a linear optimization problem to improve filter results. Experiments are performed to evaluate the walking trajectory reconstruction and the overall root mean square errors (RMSEs) of walking stride length estimation is about 3.8 cm.

Equine Kinematic Gait Analysis Using Stereo Videography and Deep Learning: Stride Length and Stance Duration Estimation

Highlights Stereo machine vision and deep learning techniques were investigated for infield equine kinematic gait analysis. The proposed pipeline tracks equine body landmarks in 3D space and estimates stride length and stance duration. The system can serve as a cost-effective, rapid, and easy-to-use tool for equine locomotion research. Abstract. Equine kinematic gait analysis (EKGA) currently requires a complicated, expensive, and labor-intensive procedure for equine locomotion research. An automated stereo video processing pipeline was developed and evaluated for measuring equine biomechanical parameters. Using stereo videos of 40 different walking horses, a DeepLabCut (DLC) model was trained to detect body landmarks in individual frames. With an autoregressive integrated moving average filter, the landmark detection had a root mean square error of 5.14 pixels and a mean absolute error of 4.87 pixels. As a case study, methods were developed to extract stride length (SL) and stance duration (SD). Individual hoof gait phase detection was achieved using a fine-tuned Faster R-CNN model and a mode filter, yielding precision and recall values of 0.83 and 0.95, respectively. The semi-global block matching (SGBM) algorithm was used to estimate depth maps, and the accuracy was assessed by comparing head length estimation with infield measurements. A Bland-Altman analysis for DLC-detected head length in combination with SGBM-based 3D reconstruction yielded a bias of -0.014 m with upper and lower limits of agreement (LoAs) of 0.03 m and -0.061 m, respectively. Furthermore, Bland-Altman analyses on SD and SL when compared to image-level manual measurements showed biases of -0.02 sec and -0.042 m, respectively. The corresponding LoAs were (0.01907 sec, -0.24 sec) for SD and (0.04 m, -0.12 m) for SL. The proposed method showed promising potential in performing EKGA in an automated, cost-effective, and rapid manner under field conditions. Keywords: 3D Reconstruction, Animal Pose Estimation, Deep Learning, Equine Kinematic Gait Analysis, Stereo Matching.

Analysis of the Recent AI for Pedestrian Navigation With Wearable Inertial Sensors

Wearable devices embedding inertial sensors enable autonomous, seamless, and low-cost pedestrian navigation. Appealing it is, the approach faces several challenges: measurement noises, different device-carrying modes, different user dynamics, and individual walking characteristics. Recent research applies Artificial Intelligence (AI) to improve inertial navigation's robustness and accuracy. Our analysis identifies 2 main categories of AI approaches depending on the inertial signals segmentation: either using human gait events (steps or strides) or fixed-length inertial data segments. A theoretical analysis of the fundamental assumptions is carried out for each category. Two state-of-the-art AI algorithms (SELDA, RoNIN), representative of each category, and a gait-driven non-AI method (SmartWalk) are evaluated in a 2,17 km long open access dataset, representative of the diversity of pedestrians' mobility surroundings (open-sky, indoors, forest, urban, parking lot). SELDA is an AI-based stride length estimation algorithm, RoNIN is an AI-based positioning method, and SmartWalk is a gait-driven non-AI positioning method. The experimental assessment shows the distinct features in each category and their limits with respect to the underlying hypotheses. On average, SELDA, RoNIN, and SmartWalk achieve 8 m, 22 m, and 17 m average positioning errors (RMSE) respectively, on six testing tracks recorded with two volunteers in various environments.

Measurement of temporal and spatial parameters of ice hockey skating using a wearable system

Ice hockey is a dynamic and competitive sport that requires a high level of neuromuscular and cardiovascular function. An objective assessment of skating helps coaches monitor athletes' performance during training sessions and matches. This study aimed to estimate the temporal and spatial parameters of skating by proposing an optimized configuration of wearable inertial measurement units (IMUs) and validating the system compared to in-lab reference systems. Ten participants were recruited to skate on a 14 m synthetic ice surface built in a motion-capture lab. Eight original event detection methods and three more adopted from gait analysis studies were implemented to detect blades-off and skate-strikes. These temporal events were detected with high accuracy and precision using skate-mounted IMUs. Also, four novel stride length estimation methods were developed to correct the estimated skaters' position using IMUs' readouts. The stride time, contact time, stride length, and stride velocity were obtained with relative errors of 3 ± 3%, 4 ± 3%, 2 ± 6%, and 2 ± 8%, respectively. This study showed that the wearable IMUs placed on skates and pelvis enables the estimation of temporal and spatial parameters of skating with high accuracy and precision, which could help coaches monitor skaters' performance in training.

Estimation of Foot Trajectory and Stride Length during Level Ground Running Using Foot-Mounted Inertial Measurement Units.

Zero-velocity assumption has been used for estimation of foot trajectory and stride length during running from the data of foot-mounted inertial measurement units (IMUs). Although the assumption provides a reasonable initialization for foot trajectory and stride length estimation, the other source of errors related to the IMU’s orientation still remains. The purpose of this study was to develop an improved foot trajectory and stride length estimation method for the level ground running based on the displacement of the foot. Seventy-nine runners performed running trials at 5 different paces and their running motions were captured using a motion capture system. The accelerations and angular velocities of left and right feet were measured with two IMUs mounted on the dorsum of each foot. In this study, foot trajectory and stride length were estimated using zero-velocity assumption with IMU data, and the orientation of IMU was estimated to calculate the mediolateral and vertical distance of the foot between two consecutive midstance events. Calculated foot trajectory and stride length were compared with motion capture data. The results show that the method used in this study can provide accurate estimation of foot trajectory and stride length for level ground running across a range of running speeds.

EL-SLE: Efficient Learning Based Stride-Length Estimation Using a Smartphone.

The pedestrian stride-length estimation is a crucial piece of personal behavior data for many smartphone applications, such as health monitoring and indoor location. The performance of the present stride-length algorithms is suitable for simple gaits and single scenes, but when applied to sophisticated gaits or heterogeneous devices, their inaccuracy varies dramatically. This paper proposes an efficient learning-based stride-length estimation model using a smartphone to obtain the correct stride length. The model uses adaptive learning to extract different elements for changing and recognition tasks, including Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) modules. The direct fusion method maps the eigenvectors to the appropriate stride length after combining the features from the learning modules. We presented an online learning module to update the model to increase the SLE model’s generalization. Extensive experiments are conducted with heterogeneous devices or users, various gaits, and switched scenarios. The results confirm that the proposed method outperforms other state-of-the-art methods and achieves an average 4.26% estimation error rate in various environments.

A Human Gait Tracking System Using Dual Foot-Mounted IMU and Multiple 2D LiDARs.

This paper proposes a human gait tracking system using a dual foot-mounted IMU and multiple 2D LiDARs. The combining system aims to overcome the disadvantages of each single sensor system (the short tracking range of the single 2D LiDAR and the drift errors of the IMU system). The LiDARs act as anchors to mitigate the errors of an inertial navigation algorithm. In our system, two 2D LiDARs are used. LiDAR 1 is placed around the starting point, and LiDAR 2 is placed at the ending point (in straight walking) or at the turning point (in rectangular path walking). Using the LiDAR 1, we can estimate the initial headings and positions of each IMU without any calibration process. We also propose a method to calibrate two LiDARs that are placed far apart. Then, the measurement from two LiDARs can be combined in a Kalman filter and the smoother algorithm to correct the two estimated feet trajectories. If straight walking is detected, we update the current stride heading and the foot position using the previous stride headings. Then, it is used as a measurement update in the Kalman filter. In the smoother algorithm, a step width constraint is used as a measurement update. We evaluate the stride length estimation through a straight walking experiment along a corridor. The root mean square errors compared with an optical tracking system are less than 3 cm. The performance of proposed method is also verified with a rectangular path walking experiment.

MHealth Technologies Toward Active Health Information Collection and Tracking in Daily Life: A Dynamic Gait Monitoring Example

Monitoring the changes in gait patterns is important to individuals’ health. Gait analysis should be taken as early as possible to prevent gait impairments and improve gait quality. Accurate stride-length estimation and gait rehabilitation activity recognition are fundamental components in gait monitoring, gait analysis, and long-term gait care. This article proposes a novel multimodality deep learning architecture to investigate the applications of stride length (SL) estimation and rehabilitation activity recognition. In order to verify this architecture, we have conducted the data collection and data labeling with our customized wearable sensing system. The sensing system can provide sensor readings from 96 sensors-based pressure array and 3-channels accelerometer and gyroscope. Many experiments with multiple perspective analysis are implemented to evaluate the models’ precision, robustness, and reliability. The multimodality deep learning architecture can map multiple sensor readings to the resulting SL with a mean absolute error of 3.89 cm and accurately detect the gait activity with an accuracy of 97.08%. It correlates the step length estimation and gait activity recognition to fulfill comprehensive long-term gait information statistic. The proposed applications’ implementation enriched our previous gait study and brought insights for clinically relevant wearable gait monitoring and gait analysis.

Adaptive Pedestrian Stride Estimation for Localization: From Multi-Gait Perspective.

Accurate and reliable stride length estimation modules play a significant role in Pedestrian Dead Reckoning (PDR) systems, but the accuracy of stride length calculation suffers from individual differences. This paper presents a stride length prediction strategy for PDR systems that can be adapted across individuals and broad walking velocity fields. It consists of a multi-gait division algorithm, which can divide a full stride into push-off, swing, heel-strike, and stance based on multi-axis IMU data. Additionally, based on the acquired gait phases, the correlation between multiple features of distinct gait phases and the stride length is analyzed, and multi regression models are merged to output the stride length value. In experimental tests, the gait segmentation algorithm provided gait phases division with the F-score of 0.811, 0.748, 0.805, and 0.819 for stance, push-off, swing, heel-strike, respectively, and IoU of 0.482, 0.69, 0.509 for push-off, swing, heel-strike, respectively. The root means square error (RMSE) of our proposed stride length estimation was 151.933, and the relative error for total distance in varying walking speed tests was less than 2%. The experimental results validated that our proposed gait phase segmentation algorithm can accurately recognize gait phases for individuals with wide walking speed ranges. With no need for parameter modification, the stride length method based on the fusion of multiple predictions from different gait phases can provide better accuracy than the estimations based on the full stride.

Mode-Independent Stride Length Estimation With IMUs in Smartphones

Robust and accurate human stride length estimation (SLE) using smartphone integrated inertial measurement units (IMU) is essential in pedestrian dead reckoning (PDR) and mobile health applications. However, the change of smartphone carrying mode (i.e. sensor location) in daily usage often leads to significant estimation errors. To address this problem, we propose a novel SLE framework called Mode-Independent Neural Network (MINN) using multi-source unsupervised domain adaptation (UDA) methods. First, we present a hierarchical neural network to extract spatio-temporal features based on multi-level ResNet and GRU. Then, we use adversarial training and a subclass classifier to build a UDA network that can extract mode-invariant features shared by the data from different modes. Finally, we integrate these architectures into an end-to-end learning framework. Through a systematic evaluation under the leave-one-out setting on two public SLE datasets, the MINN outperforms the state-of-the-art algorithms by achieving stride length error rates of 2.5% and 5.1% in supervised settings. We also evaluated the mode-independent adaptability of this model by performing single and multiple UDA tasks. The results demonstrate that the proposed MINN significantly improves the generalization of SLE model under new subjects or modes.

The Diverse Gait Dataset: Gait Segmentation Using Inertial Sensors for Pedestrian Localization with Different Genders, Heights and Walking Speeds.

Stride length estimation is one of the most crucial aspects of Pedestrian Dead Reckoning (PDR). Due to the measurement noise of inertial sensors, individual variances of pedestrians, and the uncertainty in pedestrians walking, there is a substantial error in the assessment of stride length, which causes the accumulated deviation of Pedestrian Dead Reckoning (PDR). With the help of multi-gait analysis, which decomposes strides in time and space with greater detail and accuracy, a novel and revolutionary stride estimating model or scheme could improve the performance of PDR on different users. This paper presents a diverse stride gait dataset by using inertial sensors that collect foot movement data from people of different genders, heights, and walking speeds. The dataset contains 4690 walking strides data and 19,083 gait labels. Based on the dataset, we propose a threshold-independent stride segmentation algorithm called SDATW and achieve an F-measure of 0.835. We also provide the detailed results of recognizing four gaits under different walking speeds, demonstrating the utility of our dataset for helping train stride segmentation algorithms and gait detection algorithms.

High-Performance Stacked Ensemble Model for Stride Length Estimation with Potential Application in Indoor Positioning Systems

The pedestrian dead reckoning (PDR) infrastructure-independent positioning method is of particular interest for many researchers due to its effectiveness in indoor positioning systems. The PDR technique consists of three main components including stride detection, movement heading estimation and stride length estimation (SLE). Among those, SLE results can be utilized in numerous applications, such as indoor positioning, disease diagnosis, incident detecting for patient warning, etc.. Traditional solutions for estimating stride length usually have simple structures with short calculation time but they do not ensure the expected accuracy as desired. To enhance the estimated stride length, recently, many solutions employing deep learning based techniques such as Convolutional Neural Networks (CNN), Long Short Term Memory (LSTM), have been developed to improve the SLE result. Although the accuracy in SLE has been improved compared to traditional methods, these deep learning based solutions still remain some limitations, such as complex structure and huge training parameters. This paper proposes some solutions of using machine learning based stacked ensemble model and simple machine learning algorithm to improve the accuracy of SLE while satisfying requirements such as simple structures, short execution time. The experimental results on real field data show that the proposed approaches outperform other state-of-the-art deep learning based methods on both SLE accuracy and computation time by at least 12% more accurate and seven times faster, respectively.

Image-based gramian angular field processing for pedestrian stride-length estimation using convolutional neural network

<p>In an age when people spend most of their time indoors and smartphones become a necessity, there is an increasing demand to navigate user absolute position in indoor environments. While global positioning system (GPSs) perform well outdoors, their inaccuracy can not be tolerated in places where GPS signal is weak or barely detected. This leads to a number of solutions which utilize smartphone inertial measurement unit (IMU) to track user location. Most IMU-based methods track the trajectory of a person by using stride-length and heading estimation. Thus, the accuracy of stride-length estimation plays a very important role in these methods. Inspired by recent success in the field of computer vision and machine learning, we proposed an image-based stride-length estimation method that employs gramian angular field (GAF) in converting accelerometer data into images, and then feed them into a convolutional neural network (CNN) to predict the stride-length. We evaluate the performance of our proposed method by using a public dataset from Qu Wang in his GitHub repository (available at https://github.com/Archeries/StrideLengthEstimation). The result shows that our proposed method is superior in terms of accuracy in one stride and in large walking distance than others using only data collected from the accelerometer.</p>

Inertial Sensor-Based Step Length Estimation Model by Means of Principal Component Analysis.

Inertial sensor-based step length estimation has become increasingly important with the emergence of pedestrian-dead-reckoning-based (PDR-based) indoor positioning. So far, many refined step length estimation models have been proposed to overcome the inaccuracy in estimating distance walked. Both the kinematics associated with the human body during walking and actual step lengths are rarely used in their derivation. Our paper presents a new step length estimation model that utilizes acceleration magnitude. To the best of our knowledge, we are the first to employ principal component analysis (PCA) to characterize the experimental data for the derivation of the model. These data were collected from anatomical landmarks on the human body during walking using a highly accurate optical measurement system. We evaluated the performance of the proposed model for four typical smartphone positions for long-term human walking and obtained promising results: the proposed model outperformed all acceleration-based models selected for the comparison producing an overall mean absolute stride length estimation error of 6.44 cm. The proposed model was also least affected by walking speed and smartphone position among acceleration-based models and is unaffected by smartphone orientation. Therefore, the proposed model can be used in the PDR-based indoor positioning with an important advantage that no special care regarding orientation is needed in attaching the smartphone to a particular body segment. All the sensory data acquired by smartphones that we utilized for evaluation are publicly available and include more than 10 h of walking measurements.

IMU Based Deep Stride Length Estimation With Self-Supervised Learning

Stride length estimation using inertial measurement unit (IMU) sensors is getting popular recently as one representative gait parameter for health care and sports training. The traditional estimation method requires some explicit calibrations and design assumptions. Current deep learning methods suffer from few labeled data problem. To solve above problems, this paper proposes a single convolutional neural network (CNN) model to predict stride length of running and walking and classify the running or walking type per stride. The model trains its pretext task with self-supervised learning on a large unlabeled dataset for feature learning, and its downstream task on the stride length estimation and classification tasks with supervised learning with a small labeled dataset. The proposed model can achieve better average percent error, 4.78\%, on running and walking stride length regression and 99.83\% accuracy on running and walking classification, when compared to the previous approach, 7.44\% on the stride length estimation.

Stride length and cadence estimation system in overground walking using inertial sensors

Stride length and cadence estimation system in overground walking using inertial sensors

Indoor PDR Positioning Assisted by Acoustic Source Localization, and Pedestrian Movement Behavior Recognition, Using a Dual‐Microphone Smartphone

In recent years, the public’s demand for location services has increased significantly. As outdoor positioning has matured, indoor positioning has become a focus area for researchers. Various indoor positioning methods have emerged. Pedestrian dead reckoning (PDR) has become a research hotspot since it does not require a positioning infrastructure. An integral equation is used in PDR positioning; thus, errors accumulate during long‐term operation. To eliminate the accumulated errors in PDR localisation, this paper proposes a PDR localisation system applied to complex scenarios with multiple buildings and large areas. The system is based on the pedestrian movement behavior recognition algorithm proposed in this paper, which recognises the behavior of pedestrians for each gait and improves the stride length estimation for PDR localisation based on the recognition results to reduce the accumulation of errors in the PDR localisation algorithm itself. At the same time, the system uses self‐researched hardware to modify the audio equipment used for broadcasting within the indoor environment, to locate the acoustic source through a Hamming distance‐based localisation algorithm, and to correct the estimated acoustic source estimated location based on the known source location in order to eliminate the accumulated error in PDR localisation. Through analysis and experimental verification, the recognition accuracy of pedestrian movement behavior recognition proposed in this paper reaches 95% and the acoustic source localisation accuracy of 0.32 m during movement, thus, producing an excellent effect on eliminating the cumulative error of PDR localisation.