Human Gait Recognition Based on Sequential Deep Learning and Best Features Selection

Gait recognition is an active research area that uses a walking theme to identify the subject correctly.Human Gait Recognition(HGR)is performed without any cooperation from the individual.However,in practice,it remains a challenging task under diverse walking sequences due to the covariant factors such as normal walking and walking with wearing a coat.Researchers,over the years,have worked on successfully identifying subjects using different techniques,but there is still room for improvement in accuracy due to these covariant factors.This paper proposes an automated model-free framework for human gait recognition in this article.There are a few critical steps in the proposed method.Firstly,optical flow-based motion region esti-mation and dynamic coordinates-based cropping are performed.The second step involves training a fine-tuned pre-trained MobileNetV2 model on both original and optical flow cropped frames;the training has been conducted using static hyperparameters.The third step proposed a fusion technique known as normal distribution serially fusion.In the fourth step,a better optimization algorithm is applied to select the best features,which are then classified using a Bi-Layered neural network.Three publicly available datasets,CASIA A,CASIA B,and CASIA C,were used in the experimental process and obtained average accuracies of 99.6%,91.6%,and 95.02%,respectively.The proposed framework has achieved improved accuracy compared to the other methods.

Feature Fusion Based Deep Transfer Learning Based Human Gait Classification Model

Gait is a biological typical that defines the method by that people walk.Walking is the most significant performance which keeps our day-to-day life and physical condition.Surface electromyography(sEMG)is a weak bioelectric signal that portrays the functional state between the human muscles and nervous system to any extent.Gait classifiers dependent upon sEMG signals are extremely utilized in analysing muscle diseases and as a guide path for recovery treatment.Several approaches are established in the works for gait recognition utilizing conventional and deep learning(DL)approaches.This study designs an Enhanced Artificial Algae Algorithm with Hybrid Deep Learning based Human Gait Classification(EAAA-HDLGR)technique on sEMG signals.The EAAA-HDLGR technique extracts the time domain(TD)and frequency domain(FD)features from the sEMG signals and is fused.In addition,the EAAA-HDLGR technique exploits the hybrid deep learning(HDL)model for gait recognition.At last,an EAAA-based hyperparameter optimizer is applied for the HDL model,which is mainly derived from the quasi-oppositional based learning(QOBL)concept,showing the novelty of the work.A brief classifier outcome of the EAAA-HDLGR technique is examined under diverse aspects,and the results indicate improving the EAAA-HDLGR technique.The results imply that the EAAA-HDLGR technique accomplishes improved results with the inclusion of EAAA on gait recognition.

Kinematic simulation of human gait with a multi-rigid-body foot model

The paper builds a multi-rigid-body model of human with a 4-rigid-body foot in the 3D CAD software Solidworks, based on human anatomy. By controlling the rotation of the ankle and major joints of human body while walking, the Kinematic simulation was performed in the dynamics simulation software ADAMS. The paper analyzes the simulate results and points out deficiencies in the current work and the direction of research efforts in future.

Human Gait Recognition:A Deep Learning and Best Feature Selection Framework

Background—Human Gait Recognition(HGR)is an approach based on biometric and is being widely used for surveillance.HGR is adopted by researchers for the past several decades.Several factors are there that affect the system performance such as the walking variation due to clothes,a person carrying some luggage,variations in the view angle.Proposed—In this work,a new method is introduced to overcome different problems of HGR.A hybrid method is proposed or efficient HGR using deep learning and selection of best features.Four major steps are involved in this work-preprocessing of the video frames,manipulation of the pre-trained CNN model VGG-16 for the computation of the features,removing redundant features extracted from the CNN model,and classification.In the reduction of irrelevant features Principal Score and Kurtosis based approach is proposed named PSbK.After that,the features of PSbK are fused in one materix.Finally,this fused vector is fed to the One against All Multi Support Vector Machine(OAMSVM)classifier for the final results.Results—The system is evaluated by utilizing the CASIA B database and six angles 00◦,18◦,36◦,54◦,72◦,and 90◦are used and attained the accuracy of 95.80%,96.0%,95.90%,96.20%,95.60%,and 95.50%,respectively.Conclusion—The comparison with recent methods show the proposed method work better.

Human Gait Kinematic Measurement

This study presents the measurement of the angular position and acceleration during the leg gait and the data fitting using Fourier series to parameterize the measurements obtained through accelerometers. The sensor reference is the gravity direction for the three axes that is converted into angular position and acceleration data. For this study, measurements were made in the femoral area of a human leg. The curves were obtained based on Fourier series, and though a homologation made to their harmonics, we obtained ordinary differential equations (ODEs) that parameterize these curves. The curves were approximated using six harmonics, resulting in six ODEs. The summed solutions of the ODEs represent the angular position or inclination of the leg during a walk step. The first and second derivative of the ODEs means the velocity and the acceleration of the leg movement. These types of tools are required in different research subjects such as health, entertainment, and engineering.

An Advanced Scheme of Compressed Sensing of Acceleration Data for Telemonintoring of Human Gait

The compressed sensing (CS) of acceleration data has been drawing increasing attention in gait telemonitoring application. In such application, there still exist some challenging issues including high energy consumption of body-worn device for acceleration data acquisition and the poor reconstruction performance due to nonsparsity of acceleration data. Thus, the novel scheme of compressive sensing of acceleration data is needed urgently for solutions that are found to these issues.

GaitDONet: Gait Recognition Using Deep Features Optimization and Neural Network

Human gait recognition(HGR)is the process of identifying a sub-ject(human)based on their walking pattern.Each subject is a unique walking pattern and cannot be simulated by other subjects.But,gait recognition is not easy and makes the system difficult if any object is carried by a subject,such as a bag or coat.This article proposes an automated architecture based on deep features optimization for HGR.To our knowledge,it is the first architecture in which features are fused using multiset canonical correlation analysis(MCCA).In the proposed method,original video frames are processed for all 11 selected angles of the CASIA B dataset and utilized to train two fine-tuned deep learning models such as Squeezenet and Efficientnet.Deep transfer learning was used to train both fine-tuned models on selected angles,yielding two new targeted models that were later used for feature engineering.Features are extracted from the deep layer of both fine-tuned models and fused into one vector using MCCA.An improved manta ray foraging optimization algorithm is also proposed to select the best features from the fused feature matrix and classified using a narrow neural network classifier.The experimental process was conducted on all 11 angles of the large multi-view gait dataset(CASIA B)dataset and obtained improved accuracy than the state-of-the-art techniques.Moreover,a detailed confidence interval based analysis also shows the effectiveness of the proposed architecture for HGR.

人体穿戴髋关节助力外骨骼的行走运动学分析

针对现有下肢助力外骨骼研究中缺少对关节层面的运动学分析以及缺少对髋关节外骨骼助力机制的研究这一现状,对一款髋关节助力外骨骼在多种助力模式下的运动进行采集和分析,得到被试在不穿戴外骨骼、外骨骼零助力、外骨骼低助力、外骨骼中助力、外骨骼高助力和外骨骼阻力共6种模式下的运动数据,并通过逆运动学计算和数据分析,得到关节角度曲线和步态特征。实验结果在关节层面明确了髋关节外骨骼的助力机制,可为助力外骨骼的设计和运动控制提供参考。

肌电和足压信息融合的外骨骼步态识别

为解决基于单一信号识别步态相位不够精准的问题,开展了动态交互力激励下的人机协同行走的步态识别研究。设计了肌电和足压信息采集的多模态传感器检测硬件平台;分别对单一信号开展滤波降噪、特征提取与降维等预处理;将表征下肢生理信息的肌电信号与运动信息的足压信号相融合,构建了支持向量机-模糊C均值(support vector machine-fuzzy C-mean algorithm,SVM-FCM)多模信息融合的外骨骼助行步态识别算法;开展了人机协同助行实验,实验结果表明:信息融合后的人机步态相位平均识别率达到82.49%,优于使用单一信号的识别效果,验证了多模信息融合算法识别人机协同步态的有效性。本研究可用于下肢外骨骼机器人运动控制,为人机运动相融奠定基础。

基于支持向量机的人体异常步态特征识别方法研究

人体异常步态特征识别可分析个体的行走姿势和模式,推算身份信息及人体潜在的健康问题。基于此,文章系统阐述基于支持向量机(Support Vector Machine,SVM)的人体异常步态特征识别方法,分析SVM在处理步态数据方面的技术优势和实现过程,开展CASIA-B和OUMVLP数据集的测试实验,验证该方法在步态识别上的准确性比传统反向传播(Back Propagation,BP)神经网络更高,为复杂行为识别研究提供了新视角。

下肢康复机器人机构研究现状及临床应用初探

人体下肢康复机器人作为一种恢复下肢运动能力的辅助设备,在康复领域和临床应用上的作用日益显著。随着科技的进步,国内外在该领域均有较大的发展,本研究对下肢康复机器人的发展脉络进行了较为细致地梳理,对临床应用的现状进行了综述,以机构研究为重点,从自由度,工作空间,奇异性,步态模拟,运动学仿真及人机交互等角度对其进行分析辨识,结果表明,国内对下肢康复机器人的研究重点在机构构型的设计与优化,国外的研究重点则是基于人机交互的控制系统与训练模式的提升与创新。最后结合研究现状对未来的发展趋势进行了展望。

无标记运动捕捉系统在临床步态分析上的研究进展

步态分析是临床研究的重要组成部分,传统的三维步态分析由于设备成本高、专业技术要求高等原因难以广泛应用。随着计算机视觉技术的发展,无标记运动捕捉系统的精确度得到极大提升,无需标记、成本较低的优势逐渐显现,并被应用于步态分析,但由于研究人员和临床医生缺乏对该系统的整体了解,使得该技术尚不能得到广泛推广,本综述拟对无标记运动捕捉系统在临床步态分析上的应用进展进行回顾,分析目前各项技术的优势和局限性,展望无标记运动捕捉系统在临床步态分析上的发展方向,为今后无标记运动捕捉系统在临床上的推广使用提供了研究思路。

基于关键点运动轨迹建模的步态识别

步态信息作为一个新兴的生物特征,在医疗、刑侦等方面具有广泛的应用前景.研究者已经提出了很多种步态识别方法,但普遍存在适应性不强、特征描述过于复杂或缺乏可解释性等问题.针对此问题,首先,通过改进三帧差分完成对视频图像中人体轮廓的提取;然后,基于人体轮廓图获取人体骨架模型,通过骨架模型得到所需的人体关键点位置,并对视频图像中同一关键点的位置轨迹进行曲线建模;最后依据关键点轨迹曲线模型建立一种以模型参数作为步态特征向量的步态特征描述方法,并在此基础上选取合适的分类方法进行步态识别.实验结果表明,基于关键点运动轨迹模型的步态特征表达能够很好地描述步态信息,识别率也相对较高.

Computational Models to Synthesize Human Walking

The synthesis of human walking is of great interest in biomechanics and biomimetic engineering due to its predictive capabilities and potential applications in clinical biomechanics, rehabilitation engineering and biomimetic robotics. In this paper, the various methods that have been used to synthesize humanwalking are reviewed from an engineering viewpoint. This involves a wide spectrum of approaches, from simple passive walking theories to large-scale computational models integrating the nervous, muscular and skeletal systems. These methods are roughly categorized under four headings： models inspired by the concept of a CPG （Central Pattern Generator）, methods based on the principles of control engineering, predictive gait simulation using optimisation, and models inspired by passive walking theory. The shortcomings and advantages of these methods are examined, and future directions are discussed in the context of providing insights into the neural control objectives driving gait and improving the stability of the predicted gaits. Future advancements are likely to be motivated by improved understanding of neural control strategies and the subtle complexities of the musculoskeletal system during human locomotion. It is only a matter of time before predictive gait models become a practical and valuable tool in clinical diagnosis, rehabilitation engineering and robotics.

Barefoot and High-Heeled Gait: Changes in Muscles Activation Patterns

Most women like wearing high-heeled shoes for the benefit of sensuous attractiveness and self-esteem while musculoskeletal problems and gait patterns changes are often associated. The present study aimed to identify changes during the gait stance and swing phases in some lower limb muscles. In addition, abdominal muscle was included due to its importance in dynamic trunk stability, and lack of studies on the subject. Here, we found that the use of high-heeled shoes elicited not only the increasing of the electrical activity from the muscles involved in the gait cycle, but also altered the temporal sequence of their recruitment. As practical applications, these changes may be strategies to maintain stability and minimize risks of falling, but they are often associated to diseases. Women that use high-heeled shoes for prolonged time must apply specific muscle exercises to minimize its long-term effects.

A WBAN for Human Movement Kinematics and ECG Measurements

Biomedical applications of body area networks (BANs) are evolving, where taking periodic medical readings of patients via means wireless technologies at home or in the office will aid physicians to periodically supervise the patient’s medical status without having to see the patient. Thus, one important objective of BANs is to provide the doctor with the medical readings that can be collected electronically without being in close proximity to the patient. This is done through the measurement of the patient’s physiological signals via means of wearable sensors. This paper investigates wireless BAN cooperation via actual measurements of human movement kinematics and electrocardiogram (ECG), which are believed to provide patients with easy healthcare for continuous health-monitoring. The collected information will be processed using specially designed software, which in turn will enable the patient to send a full medical chart to the physician’s electronic device. In this way, physicians will have the ability to monitor their patients more efficiently.

Interframe Variation Vector:A Novel Feature for Gait Recognition

Gait representation is an important issue in gait recognition. A simple yet efficient approach,called Interframe Variation Vector (IVV),is proposed. IVV considers the spatio-temporal motion characteristic of gait,and uses the shape variation information between successive frames to represent gait signature. Different from other features,IVV rather than condenses a gait sequence into single image resulting in spatial sequence lost; it records the whole moving process in an IVV sequence. IVV can encode whole essential features of gait and preserve all the movements of limbs. Experimental results show that the proposed gait representation has a promising recognition performance.

基于人体模型约束的步态动态识别方法

针对外骨骼机器人在人机交互、医疗康复中的人体运动步态准确识别问题,提出一种基于人体模型约束的步态动态识别方法。首先,利用AMS仿真软件建立不同运动的仿真模型,根据模型约束划分步态相位,并通过回归映射建立真实数据与仿真数据间的对应关系;然后,将柔性压力传感器采集的足底压力数据以及惯性测量单元采集的足部位移数据融合为足部运动数据,并根据动态变化结合模型约束条件动态分割运动数据,以判断步态相位;最后,搭建卷积神经网络(CNN)识别行走步态相位。实验结果表明,所提方法的行走动作步态平均识别准确率为94.58%,上、下楼梯动作的平均步态识别准确率分别为93.21%和94.64%,与未经动态分割的足底压力数据的步态识别相比,分别提高了11.34、12.19和16.03个百分点。可见,通过经动态分割的足部运动数据进行CNN识别具有较高的准确率,且适用于不同动作的步态识别。

基于高效时空图卷积的异常步态识别算法研究

为实现人体异常步态的自动识别,本研究提出一种基于高效时空图卷积神经网络的异常步态识别算法,使用两个Kinect深度相机传感器提取人体三维骨架数据,基于时空图卷积采用早期多分支融合策略,生成关节、运动和骨骼三类特征。经特征融合后,使用两个时空图卷积块作为主流网络进行训练,结合时空关节注意力机制增强模型鉴别能力。在两个公开数据集上测试分别取得了99.37%和96.10%的平均准确率,实验结果高于其他基于骨架的图卷积神经网络方法。本研究提出的高效时空图卷积网络能有效鉴别异常步态,有助于异常步态的早发现、早诊断和早治疗。

基于机器学习算法的人体步态异常状态检测与识别方法

为了提高在人体步伐识别中异常状态检测的识别率和精度,引入机器学习算法,研究人体步态异常状态检测与识别方法。根据采集到的人体步态图像序列结构,构建人体步态模型。利用所构建的模型检测人体步态,并对人体步态图像进行二值化处理,实现模型参数的优化,从中提取人体步态运动特征,实现对变化的人体步态的具体描述。结合机器学习算法中的决策树、NN分类器和KNN分类器,对人体步态中的异常状态进行识别。经过对比实验可知,本文所提识别方法具备更高的识别率和识别精度,能够有效检测出人在行走中出现的异常状态。