An Approach for Human Posture Recognition Based on the Fusion PSE-CNN-BiGRU Model

This study proposes a pose estimation-convolutional neural network-bidirectional gated recurrent unit(PSECNN-BiGRU)fusion model for human posture recognition to address low accuracy issues in abnormal posture recognition due to the loss of some feature information and the deterioration of comprehensive performance in model detection in complex home environments.Firstly,the deep convolutional network is integrated with the Mediapipe framework to extract high-precision,multi-dimensional information from the key points of the human skeleton,thereby obtaining a human posture feature set.Thereafter,a double-layer BiGRU algorithm is utilized to extract multi-layer,bidirectional temporal features from the human posture feature set,and a CNN network with an exponential linear unit(ELU)activation function is adopted to perform deep convolution of the feature map to extract the spatial feature of the human posture.Furthermore,a squeeze and excitation networks(SENet)module is introduced to adaptively learn the importance weights of each channel,enhancing the network’s focus on important features.Finally,comparative experiments are performed on available datasets,including the public human activity recognition using smartphone dataset(UCIHAR),the public human activity recognition 70 plus dataset(HAR70PLUS),and the independently developed home abnormal behavior recognition dataset(HABRD)created by the authors’team.The results show that the average accuracy of the proposed PSE-CNN-BiGRU fusion model for human posture recognition is 99.56%,89.42%,and 98.90%,respectively,which are 5.24%,5.83%,and 3.19%higher than the average accuracy of the five models proposed in the comparative literature,including CNN,GRU,and others.The F1-score for abnormal posture recognition reaches 98.84%(heartache),97.18%(fall),99.6%(bellyache),and 98.27%(climbing)on the self-builtHABRDdataset,thus verifying the effectiveness,generalization,and robustness of the proposed model in enhancing human posture recognition.

Sleep Posture Classification Using RGB and Thermal Cameras Based on Deep Learning Model

Sleep posture surveillance is crucial for patient comfort,yet current systems face difficulties in providing compre-hensive studies due to the obstruction caused by blankets.Precise posture assessment remains challenging because of the complex nature of the human body and variations in sleep patterns.Consequently,this study introduces an innovative method utilizing RGB and thermal cameras for comprehensive posture classification,thereby enhancing the analysis of body position and comfort.This method begins by capturing a dataset of sleep postures in the form of videos using RGB and thermal cameras,which depict six commonly adopted postures:supine,left log,right log,prone head,prone left,and prone right.The study involves 10 participants under two conditions:with and without blankets.Initially,the database is normalized into a video frame.The subsequent step entails training a fine-tuned,pretrained Visual Geometry Group(VGG16)and ResNet50 model.In the third phase,the extracted features are utilized for classification.The fourth step of the proposed approach employs a serial fusion technique based on the normal distribution to merge the vectors derived from both the RGB and thermal datasets.Finally,the fused vectors are passed to machine learning classifiers for final classification.The dataset,which includes human sleep postures used in this study’s experiments,achieved a 96.7%accuracy rate using the Quadratic Support Vector Machine(QSVM)without the blanket.Moreover,the Linear SVM,when utilized with a blanket,attained an accuracy of 96%.When normal distribution serial fusion was applied to the blanket features,it resulted in a remarkable average accuracy of 99%.

Multi-Branch High-Dimensional Guided Transformer-Based 3D Human Posture Estimation

The human pose paradigm is estimated using a transformer-based multi-branch multidimensional directed the three-dimensional(3D)method that takes into account self-occlusion,badly posedness,and a lack of depth data in the per-frame 3D posture estimation from two-dimensional(2D)mapping to 3D mapping.Firstly,by examining the relationship between the movements of different bones in the human body,four virtual skeletons are proposed to enhance the cyclic constraints of limb joints.Then,multiple parameters describing the skeleton are fused and projected into a high-dimensional space.Utilizing a multi-branch network,motion features between bones and overall motion features are extracted to mitigate the drift error in the estimation results.Furthermore,the estimated relative depth is projected into 3D space,and the error is calculated against real 3D data,forming a loss function along with the relative depth error.This article adopts the average joint pixel error as the primary performance metric.Compared to the benchmark approach,the estimation findings indicate an increase in average precision of 1.8 mm within the Human3.6M sample.

Fusion of Convolutional Self-Attention and Cross-Dimensional Feature Transformationfor Human Posture Estimation

Human posture estimation is a prominent research topic in the fields of human-com-puter interaction,motion recognition,and other intelligent applications.However,achieving highaccuracy in key point localization,which is crucial for intelligent applications,contradicts the lowdetection accuracy of human posture detection models in practical scenarios.To address this issue,a human pose estimation network called AT-HRNet has been proposed,which combines convolu-tional self-attention and cross-dimensional feature transformation.AT-HRNet captures significantfeature information from various regions in an adaptive manner,aggregating them through convolu-tional operations within the local receptive domain.The residual structures TripNeck and Trip-Block of the high-resolution network are designed to further refine the key point locations,wherethe attention weight is adjusted by a cross-dimensional interaction to obtain more features.To vali-date the effectiveness of this network,AT-HRNet was evaluated using the COCO2017 dataset.Theresults show that AT-HRNet outperforms HRNet by improving 3.2%in mAP,4.0%in AP75,and3.9%in AP^(M).This suggests that AT-HRNet can offer more beneficial solutions for human posture estimation.

Posture Detection of Heart Disease Using Multi-Head Attention Vision Hybrid(MHAVH)Model

Cardiovascular disease is the leading cause of death globally.This disease causes loss of heart muscles and is also responsible for the death of heart cells,sometimes damaging their functionality.A person’s life may depend on receiving timely assistance as soon as possible.Thus,minimizing the death ratio can be achieved by early detection of heart attack(HA)symptoms.In the United States alone,an estimated 610,000 people die fromheart attacks each year,accounting for one in every four fatalities.However,by identifying and reporting heart attack symptoms early on,it is possible to reduce damage and save many lives significantly.Our objective is to devise an algorithm aimed at helping individuals,particularly elderly individuals living independently,to safeguard their lives.To address these challenges,we employ deep learning techniques.We have utilized a vision transformer(ViT)to address this problem.However,it has a significant overhead cost due to its memory consumption and computational complexity because of scaling dot-product attention.Also,since transformer performance typically relies on large-scale or adequate data,adapting ViT for smaller datasets is more challenging.In response,we propose a three-in-one steam model,theMulti-Head Attention Vision Hybrid(MHAVH).Thismodel integrates a real-time posture recognition framework to identify chest pain postures indicative of heart attacks using transfer learning techniques,such as ResNet-50 and VGG-16,renowned for their robust feature extraction capabilities.By incorporatingmultiple heads into the vision transformer to generate additional metrics and enhance heart-detection capabilities,we leverage a 2019 posture-based dataset comprising RGB images,a novel creation by the author that marks the first dataset tailored for posture-based heart attack detection.Given the limited online data availability,we segmented this dataset into gender categories(male and female)and conducted testing on both segmented and original datasets.The training accuracy of our model reached an impressive 99.77%.Upon testing,the accuracy for male and female datasets was recorded at 92.87%and 75.47%,respectively.The combined dataset accuracy is 93.96%,showcasing a commendable performance overall.Our proposed approach demonstrates versatility in accommodating small and large datasets,offering promising prospects for real-world applications.

Computer Vision-Based Human Body Posture Correction System

With the development of technology and the progress of life,more and more people,regardless of entertainment,learning,or work,cannot do without computer desks and cannot put down their mobile phones.Due to prolonged sitting and often neglecting the importance of posture,incorrect posture can often lead to health problems such as hunchback,lumbar muscle strain,and shoulder and neck pain over time.To address this issue,we designed a computer vision-based human body posture detection system.The system utilizes YOLOv8 technology to accurately locate key points of the human body skeleton,and then analyzes the coordinate positions and depth information of these key points to establish a criterion for distinguishing different postures.With the assistance of an SVM classifier,the system achieves an average recognition rate of 95%.Finally,we successfully deployed the posture detection system on Raspberry Pi hardware and conducted extensive testing.The test results demonstrate that the system can effectively detect various postures and provide real-time reminders to users to correct poor posture,demonstrating good practicality and stability.

Digital Evaluation of Sitting Posture Comfort in Human-vehicle System under Industry 4.0 Framework

Most of the previous studies on the vibration ride comfort of the human-vehicle system were focused only on one or two aspects of the investigation. A hybrid approach which integrates all kinds of investigation methods in real environment and virtual environment is described. The real experimental environment includes the WBV（whole body vibration） test, questionnaires for human subjective sensation and motion capture. The virtual experimental environment includes the theoretical calculation on simplified 5-DOF human body vibration model, the vibration simulation and analysis within ADAMS/VibrationTM module, and the digital human biomechanics and occupational health analysis in Jack software. While the real experimental environment provides realistic and accurate test results, it also serves as core and validation for the virtual experimental environment. The virtual experimental environment takes full advantages of current available vibration simulation and digital human modelling software, and makes it possible to evaluate the sitting posture comfort in a human-vehicle system with various human anthropometric parameters. How this digital evaluation system for car seat comfort design is fitted in the Industry 4.0 framework is also proposed.

A Human Body Posture Recognition Algorithm Based on BP Neural Network for Wireless Body Area Networks

Human body posture recognition has attracted considerable attention in recent years in wireless body area networks(WBAN). In order to precisely recognize human body posture,many recognition algorithms have been proposed.However, the recognition rate is relatively low. In this paper, we apply back propagation(BP) neural network as a classifier to recognizing human body posture, where signals are collected from VG350 acceleration sensor and a posture signal collection system based on WBAN is designed. Human body signal vector magnitude(SVM) and tri-axial acceleration sensor data are used to describe the human body postures. We are able to recognize 4postures: Walk, Run, Squat and Sit. Our posture recognition rate is up to 91.67%. Furthermore, we find an implied relationship between hidden layer neurons and the posture recognition rate. The proposed human body posture recognition algorithm lays the foundation for the subsequent applications.

Research on Passive Locating Method Using Phase Rate of Change with Variant Posture of the Observer

Technology of passive location has broad prospects in applications. In this paper, the method using the phase rate of change for the single observer passive location is introduced based on existing methods. One can obtain the direction of the target with phase information of two orthogonal interferometers on the observer and the radial distance with the corresponding phase rate of change. Then the target can be located with high speed and precision. A locating approach is given when the flying posture of t...

Motion/Posture Modeling and Simulation Verification of Physically Handicapped in Manufacturing System Design

Non-obstacle design is critical to tailor physically handicapped workers in manufacturing system. Simultaneous consideration of variability in physically disabled users, machines and environment of the manufacturing system is extremely complex and generally requires modeling of physically handicapped interaction with the system. Most current modeling either concentrates on the task results or functional disability. The integration of physical constraints with task constraints is far more complex because of functional disability and its extended influence on adjacent body parts. A framework is proposed to integrate the two constraints and thus model the specific behavior of the physical handicapped in virtual environment generated by product specifications. Within the framework a simplified model of physical disabled body is constructed, and body motion is generated based on 3 levels of constraints（effecter constraints, kinematics constraints and physical constraints）. The kinematics and dynamic calculations are made and optimized based on the weighting manipulated by the kinematics constraints and dynamic constraints. With object transferring task as example, the model is validated in Jack 6.0. Modelled task motion elements except for squatting and overreaching well matched with captured motion elements. The proposed modeling method can model the complex behavior of the physically handicapped by integrating both task and physical disability constraints.

Design and Efficacy of Surgery for Horizontal Idiopathic Nystagmus with Abnormal Head Posture and Strabismus

The design and efficacy of surgery for horizontal idiopathic nystagmus （HIN） with abnormal head posture and strabismus were investigated. Different surgical procedures were selected according to the angle of head turn in 44 cases of HIN with abnormal head posture and strabismus. For patients with a head turn of 15° or less, the Anderson procedure was used; the yoke muscles were recessed upon slow-phase. For patients with a head turn between 15° and 25°, the surgery was designed as a Kestenbaum 5-4-4-5 procedure. For patients with a head turn of 25° or more, the surgery was designed as a Parks 5-8-6-7 procedure. The surgery to correct the abnormal head posture was performed on the fixating eye while that to correct the deviation was then performed on the non-fixating eye at the same time. The amount of surgery of the horizontal rectus muscles on the nonfixating eye was sum of the angle of head turn and the degree of deviation, which was calculated as follows： recession/resection amount of medial and lateral rectis/2×5=angle of head turn±degree of deviation. The results showed as follows： （1） Visual acuity： the visual acuity in the primary ocular position increased two lines or more in 35 patients, accounting for 79.55%. Nine patients had no or only one-line improvement, accounting for 20.45% of the entire study population; （2） The degree of deviation in the primary ocular position： 37 cases had a normal primary ocular position or the degree of deviation ≤8△ after surgery, accounting for 84.09%. Six patients had a residual degree of deviation of 8△―15△, accounting for 13.64%. One patient had a residual degree of deviation 〉20△, accounting for 2.27% of the patients examined; （3） Abnormal head posture： 34 patients had a normal head posture or a head turn of less than 5°, accounting for 72.27%. Eight patients had a residual head turn of 5°―15°, accounting for 18.18%. Two patients had a head turn of 15°― 25°, accounting for 4.55%. It was concluded that different surgical procedures based on the angle of head turn and the relationship between deviation and null zone can eliminate anomalous head posture, correct deviation, and improve vision acuity in the primary ocular position simultaneously.

Investigation of compensatory postures with videofluoromanometry in dysphagia patients

AIM: To investigate the effectiveness of head compensatory postures to ensure safe oropharyngeal transit. METHODS: A total of 321 dysphagia patients were enrolled and assessed with videofluoromanometry (VFM). The dysphagia patients were classified as follows: safe transit; penetration without aspiration; aspiration before, during or after swallowing; multiple aspirations and no transit. The patients with aspiration or no transit were tested with VFM to determine whether compensatory postures could correct their swallowing disorder. RESULTS: VFM revealed penetration without aspiration in 71 patients (22.1%); aspiration before swallowing in 17 patients (5.3%); aspiration during swallowing in 32 patients (10%); aspiration after swallowing in 21 patients (6.5%); multiple aspirations in six patients (1.9%); no transit in five patients (1.6%); and safe transit in 169 patients (52.6%). Compensatory postures guaranteed a safe transit in 66/75 (88%) patients with aspiration or no transit. A chin-down posture achieved a safe swallow in 42/75 (56%) patients, a head-turned posture in 19/75 (25.3%) and a hyperextended head posture in 5/75 (6.7%). The compensatory postures were not effective in 9/75 (12%) cases. CONCLUSION: VFM allows the speech-language therapist to choose the most effective compensatory posture without a trial-and-error process and check the effectiveness of the posture.

Gravitational effects on global hemodynamics in different postures:A closed-loop multiscale mathematical analysis

We present a novel methodology and strategy to predict pressures and flow rates in the global cardiovascular network in different postures varying from supine to upright. A closed-loop, multiscale mathematical model of the entire cardiovascular system (CVS) is developed through an integration of one-dimensional (1D) modeling of the large systemic arteries and veins, and zero-dimensional (0D) lumped-parameter modeling of the heart, the cardiac-pulmonary circulation, the cardiac and venous valves, as well as the microcirculation. A versatile junction model is proposed and incorporated into the 1D model to cope with splitting and/or merging flows across a multibranched junction, which is validated to be capable of estimating both subcritical and supercritical flows while ensuring the mass conservation and total pressure continuity. To model gravitational effects on global hemodynamics during postural change, a robust venous valve model is further established for the 1D venous flows and distributed throughout the entire venous network with consideration of its anatomically realistic numbers and locations. The present integrated model is proven to enable reasonable prediction of pressure and flow rate waveforms associated with cardiopulmonary circulation, systemic circulation in arteries and veins, as well as microcirculation within normal physiological ranges, particularly in mean venous pressures, which well match the in vivo measurements. Applications of the cardiovascular model at different postures demonstrate that gravity exerts remarkable influence on arterial and venous pressures, venous returns and cardiac outputs whereas venous pressures below the heart level show a specific correlation between central venous and hydrostatic pressures in right atrium and veins.

Posture Maintenance Control of 2-Link Object By Nonprehensile Two-Cooperative-Arm Robot Without Compensating Friction

In this paper, a method to posture maintenance control of 2-link object by nonprehensile two-cooperative-arm robot without compensating friction is proposed. In details, a mathematical model of the 2-link object is firstly built. Based on the model, stable regions for holding motion of nonprehensile two-cooperative-arm robot are obtained while the 2-link object is kept stable on the robot arms with static friction. Among the obtained stable regions, the robust pairs of orientation angles of the 2-link object are found. Under the robust orientation angles, a feedback control system is designed to control the arms to maintain the 2-link object's posture while it is being held or lifted up. Finally, experimental results are shown to verify the effectiveness of the proposed method.

STRUCTURAL PARAMETERS CALIBRATION BY POSTURE MEASUREMENT ON PARALLEL 6-DOF PLATFORM

Some approaches to measure parallel 6-degree of freedom platform's posturestatically and to calibrate the platform's actual structural parameters by measuring a series of theplatform's varying postures are studied. In the case where high posture accuracy is requiredrelatively, to obtain the platform's actual structural parameters is very important. Threedimensions measurement with 2 theodolites are used to obtain the platform's postures statically andNewton iterative method is adopted to calibrate structural parameters. Some measures taken in themeasurement and the calibration are discussed in detail. And the experiment results of theplatform's posture control before and after the calibration are given. The results show that theplatform's posture control accuracy after the calibration is improved notably.

Binocular vision and abnormal head posture in children when watching television

AIM： To determine the association between the binocular vision and an abnormal head posture（AHP）when watching television（TV） in children 7-14 y of age.·METHODS： Fifty normal children in the normal group and 52 children with an AHP when watching TV in the AHP group were tested for spherical equivalents, far and near fusional convergence（FC） and fusional divergence（FD） amplitudes, near point of convergence, far and near heterophoria, accommodative convergence/accommodation ratio and stereoacuity. The values of these tests were compared between the two groups. The independent t test was applied at a confidence level of95%.·RESULTS： The far and near FC amplitudes and far FD amplitudes were lower in the AHP group（the far FC amplitudes： break point 13.6±5.4~△, recovery point 8.7±5.4~△.The near FC amplitudes： break point 14.5 ±7.3~△, recovery point 10. 3 ± 5. 1~△. The far FD amplitudes ： break point3.9±2.7~△, recovery point 2.6±2.3~△） compared with those in the normal group（the far FC amplitudes： break point19.1 ±6.2△, recovery point 12.4 ±4.5~△. The near FC amplitudes： break point 22.3 ±8.0~△, recovery point 16.1 ±5.7~△. The far FD amplitudes： break point 7.0 ±2.1~△,recovery point 4.6 ±1.9~△）. Other tests presented no statistically significant differences.·CONCLUSION： An association between the reduced FC and FD amplitudes and the AHP in children when watching TV is proposed in the study. This kind of AHP is considered to be an anomalous manifestation which appears in a part of puerile patients of fusional vergence dysfunction.

The Impact of Footwear on Posture, Gait and Balance

Background: The objective of this study was to investigate the effects of footwear on posture and balance while walking. The types of footwear investigated in this study were open back shoes, commonly worn by hospitalized patients, and closed back shoes. Previous studies have shown that open back shoes, or slippers, increase the risk of falling (among elderly). We hypothesized that our findings would suggest that open back shoes negatively affect gait mechanics in healthy individuals. Methods: Healthy individuals (n = 12) participated in a walking test while wearing closed back shoes and open back shoes. The explanatory variables in this study were the analysis of gait, posture, and balance before and after walking. The objective variable was footwear (closed back shoes vs. open back shoes). A paired t-test was performed to detect significant differences between the two conditions. Results: Among the test items measured, we found a significant difference in minimum wide tilt angle and left-right differences in step length and intensity while walking between the conditions of closed back shoes and open back shoes. These results suggest that open back shoes could negatively impact posture and balance while walking, even in healthy subjects. Conclusion: It is imperative to improve patient awareness of the risk of falling. We believe that the inclusion of our findings in educational pamphlets and in-house notices could help improve patient awareness and more effectively prevent falls among patients.

Evaluating a school based childhood obesity intervention for posture and comfort

Background: Research shows that students who are more active throughout the day have fewer reports of body part discomfort and greater energy expenditure needed to combat childhood obesity. Many factors may contribute to the overall health of the child, including the postures that are required to complete assigned tasks at their school workstations. Decreasing sedentary behaviors in children through the use of standing desks at school has been shown to increase calorie expenditure and may be a viable approach to the energy imbalance typical of modern children. The objective of this research was to quantify and analyze sub-optimal postures and self-reported discomfort of students during the use of traditional seated and stand-biased desks to determine whether any unintended consequences of the intervention were present. Methods: A postural analysis based on the Portable Ergonomic Observation (PEO) method was used to assess the posture of 42 elementary school students as they worked at their assigned workstation (either standing or seated). Two classrooms contained stand-biased workstations (15 students) and two classrooms had traditional seated workstations (27 students). Each student was assessed three times at 10 minutes, for a total of 30 minutes of observations each. The percent of time spent in preferred versus non-preferred postures was then computed. Student body part discomfort surveys were used to assess the discomfort of students between the two groups. The relationship between type of workstation and percent time in non-preferred postures and body discomfort was examined using Wilcoxon ranksum tests and Fisher’s exact tests, respectively. The significance level was p ≤ 0.05 for all of the two-sided tests. Results: No significant difference was found between the two groups and time spent in non-preferred postures and body discomfort, children using stand-biased workstations reported less discomfort overall. Stand-biased desks presented no additional ergonomic issues, while providing increased caloric expenditure. Conclusions: A study containing a larger sample and older children that includes postural observation throughout the school day is needed for future research.

A Survey on Artificial Intelligence in Posture Recognition

Over the years,the continuous development of new technology has promoted research in the field of posture recognition and also made the application field of posture recognition have been greatly expanded.The purpose of this paper is to introduce the latest methods of posture recognition and review the various techniques and algorithms of posture recognition in recent years,such as scale-invariant feature transform,histogram of oriented gradients,support vectormachine(SVM),Gaussian mixturemodel,dynamic time warping,hiddenMarkovmodel(HMM),lightweight network,convolutional neural network(CNN).We also investigate improved methods of CNN,such as stacked hourglass networks,multi-stage pose estimation networks,convolutional posemachines,and high-resolution nets.The general process and datasets of posture recognition are analyzed and summarized,and several improved CNNmethods and threemain recognition techniques are compared.In addition,the applications of advanced neural networks in posture recognition,such as transfer learning,ensemble learning,graph neural networks,and explainable deep neural networks,are introduced.It was found that CNN has achieved great success in posture recognition and is favored by researchers.Still,a more in-depth research is needed in feature extraction,information fusion,and other aspects.Among classification methods,HMM and SVM are the most widely used,and lightweight network gradually attracts the attention of researchers.In addition,due to the lack of 3Dbenchmark data sets,data generation is a critical research direction.

Quantification of the distribution of blood flow pressure with postures

At a different angle, this study analyzed the contour chart of blood flow pressure, extreme pressure and its position to quantify DBFP in thirteen different postures with gravity considered or not (G ≠ 0 or G = 0). The aim was to determine the suitable body positions, in which the postural model of a single vessel could be simplified to two-dimensional (2D) symmetrical one while only considering such factors as posture and gravity. Computational fluid dynamic simulations were performed. Numerical results demonstrated that the DBFP showed 2D axisymmetry at ±90° and three-dimensional (3D) asymmetry at any other posture with G ≠ 0, and 2D axisymmetrical one at any posture with G = 0. Therefore, modeling a vessel as a 2D model is feasible in space and at ±90° posture on earth. In addition, the maximum pressure occurred between the inlet and the middle of the vessel, and its position variation mainly happened in the range of 0° - 15°. For a single vessel, this study provides the first theoretical evidence for cardiovascular modeling in microgravity and may help guide the researchers in designing defense devices for astronauts or patients clinically.