Abstract:Gait rehabilitation is a common method of postoperative recovery after the user sustains an injury or disability. However, traditional gait rehabilitations are usually performed under the supervision of rehabilitation specialists, which implies that the patients cannot receive adequate gait assessment anytime and anywhere. In this article, we propose GaitTracker, a novel system to remotely and continuously perform gait monitoring and analysis by three-dimensional (3D) skeletal tracking in a wearable approach. Specifically, this system consists of four Inertial Measurement Units (IMU), which are attached on the shanks and thighs of the human body. According to the measurements from these IMUs, we can obtain the motion signals of lower limbs during gait rehabilitation. By adaptively synchronizing coordinate systems of different IMUs and building the geometric model of lower limbs, the exact gait movements can be reconstructed, and gait parameters can be extracted without any prior knowledge. GaitTracker offers three key features: (1) a unified 3D skeletal model to depict the precise gait movement and parameters in 3D space, (2) a coordinate system synchronization scheme to perform space synchronization over all the IMU sensors, and (3) an automatic estimation method for the user-specific geometric parameters. In this way, GaitTracker is able to accurately perform 3D skeletal tracking of lower limbs for gait analysis, such as evaluating the gait symmetry and the gait parameters including the swing/stance time. We implemented GaitTracker and evaluated its performance in real applications. The experimental results show that, the average error for skeleton angle estimation, joint displacement estimation, and gait parameter estimation are 3∘, 2.3%, and 3%, respectively, outperforming the state of the art.

Unsupervised Domain Adaptation for Position-Independent IMU Based Gait Analysis.

Learning Gait-conditioned Bipedal Locomotion with Motor Adaptation

Global Adaptation Meets Local Generalization: Unsupervised Domain Adaptation for 3D Human Pose Estimation.

Unsupervised Domain Adaptation for Disguised-gait-based Person Identification on Micro-Doppler Signatures

MCDCD: Multi-Source Unsupervised Domain Adaptation for Abnormal Human Gait Detection

GaitSADA: Self-Aligned Domain Adaptation for mmWave Gait Recognition

GaitTracker: 3D Skeletal Tracking for Gait Analysis Based on Inertial Measurement Units

Bi-Directional Long Short-Term Memory-Based Gait Phase Recognition Method Robust to Directional Variations in Subject's Gait Progression Using Wearable Inertial Sensor

A Weight-aware-based Multi-source Unsupervised Domain Adaptation Method for Human Motion Intention Recognition

A Visual Calibration Driven Gait Analysis Model Using Wearables

Human Gait Recognition Based on Self-Adaptive Hidden Markov Model

Domain Adaptation for Inertial Measurement Unit-based Human Activity Recognition: A Survey

TraND: Transferable Neighborhood Discovery for Unsupervised Cross-Domain Gait Recognition.

Algorithm for Gait Parameters Estimation Based on Heel-Mounted Inertial Sensors

Inertial Measurement Unit-Based Real-Time Adaptive Algorithm for Human Walking Pattern and Gait Event Detection

Abnormal Gait Recognition Algorithm Based on LSTM-CNN Fusion Network

Machine Learning Based Abnormal Gait Classification with IMU Considering Joint Impairment

Real-Time Onboard Human Motion Recognition Based On Inertial Measurement Units

Adaptive Threshold based ZUPT for Single IMU enabled Wearable Pedestrian Localization

Cross-Domain Self-Supervised Complete Geometric Representation Learning for Real-Scanned Point Cloud Based Pathological Gait Analysis

Inertial Multi-Agent Systems Sensor Pedestrian Positioning Based on Foot Waist Vector Difference Constraint