Joint Locations Research Articles

A Pose-Aware Global Representation Network for Human Parsing

Many recognition tasks including image/video classification, segmentation and object detection can be improved by the integration of global information. Although global information may be better represented in some recognition tasks than the others, it is worth exploring how global information from related tasks can be effectively used to improve the performance of a target task. The task of pose estimation predicts the locations of human joints, thus providing global information about the human body. In this paper, we propose a pose-aware global representation network model (PAGRnet) that exploits global information from pose estimation to enhance feature learning in human parsing. In our PAGRnet model, a novel learning module with three integrated parts is used to learn global information. The first part generates a global joint representation, while the second part learns the relationship between the pixels and joints. By integrating the global joint representation with the pixel-joint relationship, the resulting pose-aware global representation is augmented for the parsing task. Our experimental results show competitive performance of our method on the LIP, the Pascal-person-part and the ATR datasets, with reduced computation costs in comparison to other proposals of global information fusion. We also demonstrate the advantages of our feature fusion model over concatenation, pixel-wise and channel-wise relation models.

Concentration prediction and spatial origin analysis of criteria air pollutants in Shanghai

Severe air pollution events still occur frequently in Shanghai. In order to predict when Shanghai air quality satisfies the National Ambient Air Quality Standards of China (NAAQSC) and identify potential source areas of criteria air pollutants for the regional joint prevention and control of air pollution, concentration data of PM2.5, PM10, SO2, NO2 and O3 were collected in 2014–2022 at fourteen monitoring sites across Shanghai and surrounding areas. A first – order rate equation with harmonic regression analysis was employed for time series analysis and concentration prediction. Decreasing concentrations were observed widely over all sites except O3 and NO2. It is very likely that the secondary NAAQSC standards for PMx, and SO2 would be met by 2025 and O3 and NO2 would likely become the critical pollutants that determine air quality level after 2025. Regional transport was predominant for PMx and SO2 pollution. A 3D – CWT multisite joint location method was developed to identify their potential source areas at different spatial resolutions. Weighting function correction was assigned via information entropy of endpoint numbers in each cell. A probabilistic parameter WIPSA was proposed to quantify and normalize the probability that grid cells are source areas in order to achieve fourteen – site joint location, and it was comparable and compatible at different spatial resolutions. Potential source areas of PM2.5 and PM10 were similar, including Henan, Shandong, Hebei and Anhui, while origin domains of SO2 mainly covered Henan and Hebei. In all seasons, air pollution that was transported to Shanghai (i.e., PMx and SO2) originated mainly from the North China Plain; the contribution of marine sources was neglectable.

Investigation of microstructure characteristics and work hardening behaviour of water-cooled FSW dissimilar aluminium alloys

The present study investigates the microstructure characteristics, tensile properties and work hardening behavior of friction stir welding (FSW) AA5083-AA6061 dissimilar aluminium alloys under different cooling conditions (naturally air-cooled and water-cooled). Microstructure evolution and tensile test of FSWed joints were investigated through Optical Microscope (OM), Electron Back Scattered Diffraction (EBSD), Scanning Electron Microscope (SEM), X-Ray Diffraction (XRD) and Universal Testing Machine (UTM). EBSD results showed that the grain size obtained at the stir zone was 7.6 µm for air-cooled FSW (ACFSW) and grain growth restricted to 3.2 µm using water-cooled FSW (WCFSW). Brass [{110} <112>] and Goss [{110} <001>] were the two texture components that primarily evolved in FSW samples. The study suggests that joints produced with water cooling exhibited higher UTS at all distinct microstructure locations of the FSW joint and whereas the WCFSW-L sample exhibited higher UTS (271 MPa) than compared to ACFSW-L (246 MPa). Moreover, in the WCFSW, the strain hardening coefficient and work hardening rate enhanced by 6.3 % and 18.7 %, respectively, than ACFSW. The fracture surface morphology analysis of ACFSW and WCFSW samples was consistent with the tensile strength results.

TriHorn-Net: A model for accurate depth-based 3D hand pose estimation

3D hand pose estimation methods have made significant progress recently. However, estimation accuracy is often far from sufficient for specific real-world applications, and thus there is significant room for improvement. This paper proposes TriHorn-Net, a novel model that uses specific innovations to improve hand pose estimation accuracy on depth images. The first innovation is decomposition of the 3D hand pose estimation into the estimation of 2D joint locations in the depth image space (UV), and the estimation of their corresponding depths aided by two complementary attention maps. This decomposition prevents depth estimation, which is a more difficult task, from interfering with the UV estimations at both the prediction and feature levels. The second innovation is PixDropout, which is, to the best of our knowledge, the first appearance-based data augmentation method for hand depth images. Experimental results demonstrate that the proposed model outperforms the state-of-the-art methods on three public benchmark datasets, achieving 5.73 mm, 7.13 mm, and 7.68 mm on the ICVL, MSRA, and NYU datasets, respectively. The proposed model achieves this performance with a relatively fast average inference speed of 9.25 ms per frame on an NVIDIA 1080Ti GPU, which, as discussed in the paper, places the proposed model among the fastest-performing 3D hand pose estimation models. Our implementation is available at https://github.com/mrezaei92/TriHorn-Net.

Scaling Up Electric-Vehicle Battery Swapping Services in Cities: A Joint Location and Repairable-Inventory Model

Battery swapping for electric vehicle refueling is reviving and thriving. Despite a captivating sustainable future where swapping batteries will be as convenient as refueling gas today, a tension is mounting in practice (beyond the traditional “range anxiety” issue): On one hand, it is desirable to maximize battery proximity and availability to customers. On the other hand, capacitated urban power grids may curb decentralized charging at a slow speed. To reconcile this tension, some cities are embracing an emerging infrastructure network: Decentralized swapping stations replenish charged batteries from centralized charging stations. It remains unclear how to design such a network or whether pooling charging demands will save costs or batteries. In this paper, we model this new urban infrastructure network. This task is complicated by non-Poisson swaps and by the intertwined stochastic operations of swapping, charging, stocking, and circulating batteries among swapping and charging stations. We tackle these complexities by deriving analytical models, which enrich the classical batched repairable-inventory theory. We next propose a joint location and repairable-inventory model for citywide deployment of hub charging stations, with a nonconvex nonconcave objective function. We solve this problem exactly by exploiting submodularity and combining constraint-generation and parameter-search techniques. Even for solving convexified problems, our algorithm brings a speedup of at least three orders of magnitude relative to the Gurobi solver. The major insight is twofold: The benefit of pooling charging demands alone is not enough to justify the adoption of the “swap-locally, charge-centrally” network; instead, the main justification is that faster charging accessible at centralized charging stations significantly reduces the system-wide battery stock level. In a broader sense, this work deepens our understanding of how mobility and energy are coupled toward enabling smart cities. This paper was accepted by Chung Piaw Teo, optimization. Funding: Y. Zhang acknowledges the support from the National Natural Science Foundation of China [Grants 71871023, 72271029, and 72061127001]. W. Qi acknowledges the support from the National Natural Science Foundation of China [Grants 72272014 and 72188101] and the Natural Sciences and Engineering Research Council of Canada [Grant RGPIN-2019-04769]. N. Zhang acknowledges the support from the China Scholarship Council [202106030140]. Supplemental Material: The data files and online appendices are available at https://doi.org/10.1287/mnsc.2023.4731 .

Cloud Based VPN Using IP Tunneling for Remote Site Interface

Abstract: A strong IP technology that builds a secure and encrypted connection tunnel over the less secure internet is required by the majority of enterprise organizations. Joint location and link connectivity issues must be resolved for distant users and branch offices to have secure access to corporate applications and other resources. This ensures security while data passes over encrypted tunnels. Despite the development of IP-based VPN, JLP and LCP for VPN customer edge devices and provider edge nodes have not yet been completely investigated. The purpose of this work is to apply cutting-edge viewpoints to VPN-IP design that may be used in IP multi-protocol label switching (MPLSVPN) infrastructure for cloud computing. The system architecture includes mathematical formulas. End-to-end delay, throughput, and resource consumption behavior for IP tunneling are seen to behave moderately for low scale workloads. While showcasing the benefits of an MPLS-based IP-based VPN architecture, the difficulties of VPN-IP tunneling are explored. This study suggests cloud-based MPLS-VPN as a workable substitute for traditional VPN-IP tunneling in order to get the best performance and service delivery

A multi-modal fundus image based auxiliary location method of lesion boundary for guiding the layout of laser spot in central serous chorioretinopathy therapy

The lesion boundary of central serous chorioretinopathy (CSCR) is the guarantee to guide the ophthalmologist to accurately arrange the laser spots, so as to enable this ophthalmopathy to be treated precisely. Currently, the accuracy and rapidity of manually locating CSCR lesion boundary in clinic based on single-modal fundus image are limited by imaging quality and ophthalmologist experience, which is also accompanied by poor repeatability, weak reliability and low efficiency. Consequently, a multi-modal fundus image-based lesion boundary auxiliary location method is developed. Firstly, the initial location module (ILM) is employed to achieve the preliminary location of key boundary points of CSCR lesion area on the optical coherence tomography (OCT) B-scan image, then followed by the joint location module (JLM) created based on reinforcement learning for further enhancing the location accuracy. Secondly, the scanning line detection module (SLDM) is constructed to realize the location of lesion scanning line on the scanning laser ophthalmoscope (SLO) image, so as to facilitate the cross-modal mapping of key boundary points. Finally, a simple yet effective lesion boundary location module (LBLM) is designed to assist the automatic cross-modal mapping of key boundary points and enable the final location of lesion boundary. Extensive experiments show that each module can perform well on its corresponding sub task, such as JLM, which makes the correction rate (CR) of ILM increase to 92.11%, comprehensively indicating the effectiveness and feasibility of this method in providing effective lesion boundary guidance for assisting ophthalmologists to precisely arrange the laser spots, and also opening a new research idea for the automatic location of lesion boundary of other fundus diseases.

Evaluation of Spliced UHM CFRP Strip Panels for Strengthening Steel Beams

Short ultrahigh-modulus (UHM) carbon fiber–reinforced polymer (CFRP) strip panels connected through a finger joint configuration offer a convenient modular method for strengthening steel girders. Three-dimensional nonlinear finite-element (FE) analysis was conducted to characterize the interfacial and flexural behavior of strip panels. The calibrated FE model provides an efficient and economical method of evaluating different parameters related to the material properties and strengthening configuration of the UHM CFRP strip panels, for which experimental data are not available. Analysis implemented robust features, including debonding of CFRP using a mixed-mode bond–slip model, CFRP rupture, steel yielding, and concrete slab crushing. Predictions were calibrated and validated against experimental results from double lap shear tests and flexural tests of steel beams and steel–concrete composite beams. The FE model was used to investigate the effects of finger joint location, joint length, and laminate thickness. Strip panels with 300-mm-long finger joints and thicknesses of up to 2 mm equaled the strength of a continuous laminate with the same CFRP area. Both failed by CFRP rupture. At greater thicknesses, debonding occurred at lower ultimate loads than for a continuous laminate. The finger joint splice was also compared to the traditional splice plate method. The splice plate length required to achieve the same level of load transfer was longer than the finger joint length when the connection was in the maximum moment region. For a given CFRP thickness and splice plate length (or finger joint length), strip panels with finger joints provide greater load-carrying capacity.

Behaviour and load transfer mechanism of a novel I-beam to box-column connection

The recent studies on the connection of I-beam to box-column with a stub beam proposed by authors demonstrated the connection satisfactory performance. This paper discusses the load transfer mechanism of the connection. For this reason, reversed cyclic tests were conducted on the three configurations of the connection with different beam to column width ratios of 0.50, 0.67, and 0.75. The tested specimens revealed stable hysteresis curves. Furthermore, the experimental results show that the connection is fully restrained and qualified for use in special moment frames (SMF) according to AISC. The plastic hinge developed on the beam near to beam end-plate and met the requirements of extended end-plate connections. Also, the panel zone of connection remained elastic throughout the tests. The study of load transfer mechanism revealed that by transferring the major portion of load to the column webs, the normal stress in the column flange is reduced, leading to the elimination of the need for continuity plates. Moreover, the finite element models of tested specimens were simulated. After the verification of numerical models against experimental results, they were extended to examine the effects of stub beam length and beam to column width ratio on the load transfer mechanism. According to the numerical results, the load transfer mechanism at the stub beam to column joint location was not sensitive to the changes in stub beam length and beam to column width ratio.

Evaluating current state of monocular 3D pose models for golf

Monocular 3D human pose estimation has reached an impressive performance. State-of-the-art mod- els predict joint locations that can be accurately reprojected back into the image, resulting in vi- sually convincing detections. However, our aim is to use the predicted poses in a domain with high- frequency movements, that is, for video of ath- letes performing golf swings. Our investigation is based on accurate marker-based motion capture data. Also, for our data, the predicted 3D joint locations look convincing when we reproject them into the image. However, by quantitatively com- paring the results with the motion capture data, we see significant model errors that are too erroneous to be used for any kinematic analysis of the move- ments. Thus we conclude that the current models cannot be used out of the box for advanced golf analytics.

CSI-Based Joint Location and Activity Monitoring for COVID-19 Quarantine Environments

The recent SARS-COV-2 virus, also known as COVID-19, badly affected the world’s healthcare system due to limited medical resources for a large number of infected human beings. Quarantine helps in breaking the spread of the virus for such communicable diseases. This work proposes a nonwearable/contactless system for human location and activity recognition using ubiquitous wireless signals. The proposed method utilizes the channel state information (CSI) of the wireless signals recorded through a low-cost device for estimating the location and activity of the person under quarantine. We propose to utilize a Siamese architecture with combined one-dimensional convolutional neural networks (1-D-CNNs) and bi-directional long short-term memory (Bi-LSTM) networks. The proposed method provides high accuracy for the joint task and is validated on two real-world testbeds, first, using the designed low-cost CSI recording hardware, and second, on a public dataset for joint activity and location estimation. The human activity recognition (HAR) results outperform state-of-the-art machine and deep learning methods, and localization results are comparable with the existing methods.

Willow Catkin Optimization Algorithm Applied in the TDOA-FDOA Joint Location Problem

The heuristic optimization algorithm is a popular optimization method for solving optimization problems. A novel meta-heuristic algorithm was proposed in this paper, which is called the Willow Catkin Optimization (WCO) algorithm. It mainly consists of two processes: spreading seeds and aggregating seeds. In the first process, WCO tries to make the seeds explore the solution space to find the local optimal solutions. In the second process, it works to develop each optimal local solution and find the optimal global solution. In the experimental section, the performance of WCO is tested with 30 test functions from CEC 2017. WCO was applied in the Time Difference of Arrival and Frequency Difference of Arrival (TDOA-FDOA) co-localization problem of moving nodes in Wireless Sensor Networks (WSNs). Experimental results show the performance and applicability of the WCO algorithm.

A novel approach for yoga pose estimation based on in-depth analysis of human body joint detection accuracy.

Virtual motion and pose from images and video can be estimated by detecting body joints and their interconnection. The human body has diverse and complicated poses in yoga, making its classification challenging. This study estimates yoga poses from the images using a neural network. Five different yoga poses, viz. downdog, tree, plank, warrior2, and goddess in the form of RGB images are used as the target inputs. The BlazePose model was used to localize the body joints of the yoga poses. It detected a maximum of 33 body joints, referred to as keypoints, covering almost all the body parts. Keypoints achieved from the model are considered as predicted joint locations. True keypoints, as the ground truth body joint for individual yoga poses, are identified manually using the open source image annotation tool named Makesense AI. A detailed analysis of the body joint detection accuracy is proposed in the form of percentage of corrected keypoints (PCK) and percentage of detected joints (PDJ) for individual body parts and individual body joints, respectively. An algorithm is designed to measure PCK and PDJ in which the distance between the predicted joint location and true joint location is calculated. The experiment evaluation suggests that the adopted model obtained 93.9% PCK for the goddess pose. The maximum PCK achieved for the goddess pose, i.e., 93.9%, PDJ evaluation was carried out in the staggering mode where maximum PDJ is obtained as 90% to 100% for almost all the body joints.

Deep learning-based real-time 3D human pose estimation

Human body pose estimation represented by joint rotations is essential for driving the virtual characters. The present paper developed a novel end-to-end point-to-pose mesh fitting network (P2P-MeshNet) to directly estimate the body joint rotations. P2P-MeshNet provided a strong collaboration between the deep learning network, an inverse kinematics network for body pose estimation (IKNet-body), and the self-correcting network, an iterative error feedback network (IEF). The introduced P2P-MeshNet was then applied to the free mocap (FreeMocap) dataset covering OpenPose 3D joint locations reconstructed from multi-view OpenPose 2D joint locations. The generated joint rotations were tested using the mean per joint position error (MPJPE), as well as the percentage of correct keypoints (PCK) along with the area under the PCK curve (AUC) with a threshold range of 0–60 mm after Procrustes aligned. Based on the compared metrics, P2P-MeshNet with 11.31 mm and 99.7% in estimate error and success rate as well as an AUC of 80.9 demonstrated a more consistent tool for future human body pose estimation. The runtime performance of 100 frames per second implied its potential application prospects.

Lateral epicondyle to the joint line distance is a precise landmark for determination of an accurate knee joint line: an observational retrospective study

PurposeTo assess a quantitative and reproducible association between the position of the knee joint line and recognizable anatomical landmarks around it in order to help in restoring joint line in arthroplasty cases.MethodsMagnetic resonance imaging (MRI) of 130 normal knees were investigated. Anatomical measurements of the knee joint distances on the obtained planes were performed manually by distance measurements using a ruler tool, followed by 6 anatomical bony landmarks determination about the knee to identify the joint line which included the joint line, medial epicondyle, lateral epicondyle, medial flare, lateral flare, and proximal tibiofibular joint. The entire process was examined twice by two independent fellowship trained musculoskeletal radiologists, with a 2-week interval between the first and second sets of readings.ResultsThe lateral epicondyle to the joint line of the knee (LEJL) could be a reliable landmark for accurate distance measurements for the knee joint line level, with an absolute distance of 24.4 ± 2.8 mm. The analysis showed that the femorotibial ratio between the LEJL and proximal tibiofibular joint (PTFJ) was 1.0 (LEJL/PTFJJL = 1.0 ± 0.1), confirming the location of the knee joint at the midpoint between the lateral epicondyle and PTFJ, revealing two identifiable landmarks.ConclusionsLEJL is the most precise landmark for determination of an accurate knee joint line because the knee is located at the midline between the lateral epicondyle and PTFJ. These reproducible quantitative relationships can be widely employed in various imaging modalities to help restore the knee JL in arthroplasty surgeries.

Relationship between carpal bone morphology and scaphoid fracture location

Aim: Morphological variations in carpal bone may have an effect on wrist pathologies. In this study, we examined the relationship between carpal bone morphologies and scaphoid fractures. Materials and Methods: Patients with a prominent scaphoid fracture on posteroanterior and ulnar deviation X-ray were included in the study. Morphological types of lunate, hamate, capitate, presence of capitate-fourth metacarpal joint and scaphoid fracture location were documented retrospectively. A total of 68 patients, 65 (95.6%) male and 3 (4.4%) female, were included in the study. Results: Considering the scaphoid fracture types, 40 (58.8%) patients had waist fractures and 28 (41.2%) patients had proximal pole fractures. The association of fracture type and morphology of carpal bones was examined. Conclusion: In our study, we found that lunate, capitate, hamatum morphology and ulnar variance, fourth metacarpal articulation had no effect on the scaphoid fracture pattern.

3D Human Pose Estimation Using Two-Stream Architecture with Joint Training

With the advancement of image sensing technology, estimating 3D human pose from monocular video has become a hot research topic in computer vision. 3D human pose estimation is an essential prerequisite for subsequent action analysis and understanding. It empowers a wide spectrum of potential applications in various areas, such as intelligent transportation, human-computer interaction, and medical rehabilitation. Currently, some methods for 3D human pose estimation in monocular video employ temporal convolutional network (TCN) to extract inter-frame feature relationships, but the majority of them suffer from insufficient inter-frame feature relationship extractions. In this paper, we decompose the 3D joint location regression into the bone direction and length, we propose the TCG, a temporal convolutional network incorporating Gaussian error linear units (GELU), to solve bone direction. It enables more inter-frame features to be captured and makes the utmost of the feature relationships between data. Furthermore, we adopt kinematic structural information to solve bone length enhancing the use of intra-frame joint features. Finally, we design a loss function for joint training of the bone direction estimation network with the bone length estimation network. The proposed method has extensively experimented on the public benchmark dataset Human3.6M. Both quantitative and qualitative experimental results showed that the proposed method can achieve more accurate 3D human pose estimations.

Exploiting Temporal Contexts With Strided Transformer for 3D Human Pose Estimation

Despite the great progress in 3D human pose estimation from videos, it is still an open problem to take full advantage of a redundant 2D pose sequence to learn representative representations for generating one 3D pose. To this end, we propose an improved Transformer-based architecture, called Strided Transformer, which simply and effectively lifts a long sequence of 2D joint locations to a single 3D pose. Specifically, a Vanilla Transformer Encoder (VTE) is adopted to model long-range dependencies of 2D pose sequences. To reduce the redundancy of the sequence, fully-connected layers in the feed-forward network of VTE are replaced with strided convolutions to progressively shrink the sequence length and aggregate information from local contexts. The modified VTE is termed as Strided Transformer Encoder (STE), which is built upon the outputs of VTE. STE not only effectively aggregates long-range information to a single-vector representation in a hierarchical global and local fashion, but also significantly reduces the computation cost. Furthermore, a full-to-single supervision scheme is designed at both full sequence and single target frame scales applied to the outputs of VTE and STE, respectively. This scheme imposes extra temporal smoothness constraints in conjunction with the single target frame supervision, improving the capability of producing smoother and more accurate 3D poses for the target frame. The proposed Strided Transformer is evaluated on two challenging benchmark datasets, Human3.6M and HumanEva-I, and achieves state-of-the-art results with fewer parameters. Code and models are available at \url{https://github.com/Vegetebird/StridedTransformer-Pose3D}.

A Joint Location–Allocation–Inventory Spare Part Optimization Model for Base-Level Support System with Uncertain Demands

This paper copes with a joint Location-Allocation-Inventory problem in a three-echelon base-level spare part support system with epistemic uncertainty in uncertain demands of bases. The aim of the paper is to propose an optimization model under the uncertainty theory to minimize the total cost, which integrates crucial characterizations of the inventory control decisions and the location-allocation scheme arrangement under a periodic review order-up-to-S (T, S) policy. Uncertainty theory is introduced in this paper to characterize epistemic uncertainty, where demands are treated as uncertain variables and stockout loss is represented by value-at-risk in uncertain measurement. To solve the original uncertain optimization model, an equivalent deterministic model is derived and addressed by an improved bilevel genetic algorithm. Moreover, the proposed models and algorithm are encoded into numerical examples for supply chain programming. The results highlight the applicability of the model and the algorithm’s effectiveness in approaching the optimal solution compared with traditional genetic algorithm. Sensitivity analyses are further made for the impacts of review time and inventory capacity on different cost components.

Optimization of Charging/Battery-Swap Station Location of Electric Vehicles with an Improved Genetic Algorithm-Based Model

The joint location planning of charging/battery-swap facilities for electric vehicles is a complex problem. Considering the differences between these two modes of power replenishment, we constructed a joint location-planning model to minimize construction and operation costs, user costs, and user satisfaction-related penalty costs. We designed an improved genetic algorithm that changes the crossover rate using the fitness value, memorizes, and transfers excellent genes. In addition, the present model addresses the problem of “premature convergence” in conventional genetic algorithms. A simulated example revealed that our proposed model could provide a basis for optimized location planning of charging/battery-swapping facilities at different levels under different charging modes with an improved computing efficiency. The example also proved that meeting more demand for power supply of electric vehicles does not necessarily mean increasing the sites of charging/battery-swap stations. Instead, optimizing the level and location planning of charging/battery-swap stations can maximize the investment profit. The proposed model can provide a reference for the government and enterprises to better plan the location of charging/battery-swap facilities. Hence, it is of both theoretical and practical value.