Geometric Network Research Articles

The effect of discharge on head loss with straight and bend flow directions in the pipeline

Piping flow is a way of transportation to carry material from one place to another. A problem in pipelines is a head loss. This head loss can affect pipe capacity and reduce flow discharge. Structure or geometry of pipelines will affect energy due to the roughness of pipe walls, friction, bends, joints and energy decreases with the length of passed pipe. Head loss in the piping system includes major and minor energy losses. The objective of this research is to determine the magnitude of major and minor head losses and to determine the relationship between discharge and velocity, pressure height, and the influence of geometric network structure. The variation of flowed discharge is 0.5 liter/second, 2.5 liter/second, 3 liter/second and 4 liter/second. The research is experimental research conducted in the laboratory. The test design uses a pipeline with a length of 6 m, a width of 1.20 m, and a dimension of 64 mm. The results showed that an increase in flow rate (Q) followed by an increase in flow velocity (v) will increase Reynold’s (Re) number so that the head loss in the pipeline will increase. The increasing of Reynold’s (Re) number will decrease the value of the friction coefficient. Structure or geometric of the line also influences the pattern of flow. In a straight pipe, the velocity of flow is large but the pressure height is small compared to the bend pipe flow where the velocity of flow is small but the pressure height is larger.

Geometric rectification‐based neural network architecture for image manipulation detection

Determination of image authenticity usually requires the identification and localization of the manipulated regions of images. Hence, image manipulation detection has become one of the most important tasks in the field of multimedia forensics. Recently, Convolutional Neural Networks (CNNs) have achieved promising performance in image manipulation detection. However, it is hard for the existing CNN-based manipulation detection approaches to accurately identify and localize the manipulated regions that have undergone geometric transformations, since CNNs are limited by their inability to be geometrically invariant. To address this issue, we propose a geometric rectification-based neural network architecture for image manipulation detection. In this type of network architecture, following the detection of a set of potential manipulated regions (PMRs) using Region Proposal Network, the Spatial Transformer Network is employed to geometrically rectify the convolutional feature maps (CFMs) of these regions to obtain the geometrically rectified CFMs (GR-CFMs). Subsequently, the residual feature maps (RFMs) are computed to capture the characteristic inconsistency between the CFMs and GR-CFMs of each PMR. Finally, the computed RFMs are automatically integrated with the GR-CFMs by a designed attention module to determine whether each PMR is a manipulated region and to localize the manipulated part at the pixel-level. Extensive experiments on the public data set as well as on our challenging data set demonstrate that the proposed network architecture achieves desirable performance in identifying and localizing regions with common tampering artifacts, which involve geometric transformations.

Automated landmarking for palatal shape analysis using geometric deep learning.

To develop and evaluate a geometric deep-learning network to automatically place seven palatal landmarks on digitized maxillary dental casts. The sample comprised individuals with permanent dentition of various ethnicities. The network was trained from manual landmark annotations on 732 dental casts and evaluated on 104 dental casts. A geometric deep-learning network was developed to hierarchically learn features from point-clouds representing the 3D surface of each cast. These features predict the locations of seven palatal landmarks. Repeat-measurement reliability was <0.3mm for all landmarks on all casts. Accuracy is promising. The proportion of test subjects with errors less than 2mm was between 0.93 and 0.68, depending on the landmark. Unusually shaped and large palates generate the highest errors. There was no evidence for a difference in mean palatal shape estimated from manual compared to the automatic landmarking. The automatic landmarking reduces sample variation around the mean and reduces measurements of palatal size. The automatic landmarking method shows excellent repeatability and promising accuracy, which can streamline patient assessment and research studies. However, landmark indications should be subject to visual quality control.

The geometric attention-aware network for lane detection in complex road scenes.

Lane detection in complex road scenes is still a challenging task due to poor lighting conditions, interference of irrelevant road markings or signs, etc. To solve the problem of lane detection in the various complex road scenes, we proposed a geometric attention-aware network (GAAN) for lane detection. The proposed GAAN adopted a multi-task branch architecture, and used the attention information propagation (AIP) module to perform communication between branches, then the geometric attention-aware (GAA) module was used to complete feature fusion. In order to verify the lane detection effect of the proposed model in this paper, the experiments were conducted on the CULane dataset, TuSimple dataset, and BDD100K dataset. The experimental results show that our method performs well compared with the current excellent lane line detection networks.

Target-based automated matching of multiple terrestrial laser scans for complex forest scenes

Terrestrial laser scanners are widely used to derive unbiased and non-destructive estimates of the vertical distribution of the plant area index and plant area volume density at plot-level scales, as well as the above-ground biomass, height, and diameter at breast height of individual trees. Multiple scans are often employed to capture and register data so that all of the stems can be detected and their complete forms can be analyzed. Researchers have traditionally preferred target-less strategies to register scans because of their low cost and convenience. However, in complex forest scenes, even state-of-the-art approaches cannot guarantee the success of any pairwise registration. In this study, we present an automated target-based processing approach for the registration of unordered scans in complex forest scenes. In contrast to previous studies, the proposed registration method automatically detects the artificial targets and builds a geometric network to judge their connectivity. A pose graph is then exploited to combine these data with the corresponding pairwise transformation, and then the scans are integrated into a unified coordinate system. This method is more robust and efficient than target-less approaches because it is independent of the characteristics of individual trees and does not require ground information. In an experimental scenario, we use an extremely complex wild bamboo forest scene to evaluate the performance of the proposed approach in terms of robustness, accuracy, and efficiency.

Hierarchical effects facilitate spreading processes on synthetic and empirical multilayer networks.

In this paper we consider the effects of corporate hierarchies on innovation spread across multilayer networks, modeled by an elaborated SIR framework. We show that the addition of management layers can significantly improve spreading processes on both random geometric graphs and empirical corporate networks. Additionally, we show that utilizing a more centralized working relationship network rather than a strict administrative network further increases overall innovation reach. In fact, this more centralized structure in conjunction with management layers is essential to both reaching a plurality of nodes and creating a stable adopted community in the long time horizon. Further, we show that the selection of seed nodes affects the final stability of the adopted community, and while the most influential nodes often produce the highest peak adoption, this is not always the case. In some circumstances, seeding nodes near but not in the highest positions in the graph produces larger peak adoption and more stable long-time adoption.

GEOMETRIC SCATTERING ATTENTION NETWORKS.

Geometric scattering has recently gained recognition in graph representation learning, and recent work has shown that integrating scattering features in graph convolution networks (GCNs) can alleviate the typical oversmoothing of features in node representation learning. However, scattering often relies on handcrafted design, requiring careful selection of frequency bands via a cascade of wavelet transforms, as well as an effective weight sharing scheme to combine low- and band-pass information. Here, we introduce a new attention-based architecture to produce adaptive task-driven node representations by implicitly learning node-wise weights for combining multiple scattering and GCN channels in the network. We show the resulting geometric scattering attention network (GSAN) outperforms previous networks in semi-supervised node classification, while also enabling a spectral study of extracted information by examining node-wise attention weights.

2D–3D Geometric Fusion network using Multi-Neighbourhood Graph Convolution for RGB-D indoor scene classification

Multi-modal fusion has been proved to help enhance the performance of scene classification tasks. This paper presents a 2D–3D Fusion stage that combines 3D Geometric Features with 2D Texture Features obtained by 2D Convolutional Neural Networks. To get a robust 3D Geometric embedding, a network that uses two novel layers is proposed. The first layer, Multi-Neighbourhood Graph Convolution, aims to learn a more robust geometric descriptor of the scene combining two different neighbourhoods: one in the Euclidean space and the other in the Feature space. The second proposed layer, Nearest Voxel Pooling, improves the performance of the well-known Voxel Pooling. Experimental results, using NYU-Depth-V2 and SUN RGB-D datasets, show that the proposed method outperforms the current state-of-the-art in RGB-D indoor scene classification task.

Identification of wool and cashmere fibers based on multiscale geometric analysis

The identification of wool and cashmere fibers has always been an essential topic in the field of textile. In this work, a fiber recognition algorithm based on multiscale geometric analysis and deep convolutional neural network is introduced. As a new tool for signal analysis with multiscale and multi-direction, curvelet transform not only has multiscale and multi-resolution characteristics of the wavelet transform, but also has very strong directionality in the sense that it can provide an optimally sparse representation of fiber image with a large number of fiber contours information. The proposed method was based on multiscale geometric analysis to reduce the dimension of wool/cashmere fiber images and reduce the calculation of redundant data. The deep convolutional neural network was used to classify and recognize wool/cashmere fiber images. Then 400 fiber images of two kinds of fiber samples including wool and cashmere fibers, were collected respectively and processed by the methods including random interception and rotation to obtain a total of 600 fiber images respectively for the experiment analysis. The results show that the recognition accuracy is up to 96.67%. Compared with the traditional feature extraction and classification algorithm, this method dramatically improves model performance and identification accuracy.

GraphRegNet: Deep Graph Regularisation Networks on Sparse Keypoints for Dense Registration of 3D Lung CTs.

In the last two years learning-based methods have started to show encouraging results in different supervised and unsupervised medical image registration tasks. Deep neural networks enable (near) real time applications through fast inference times and have tremendous potential for increased registration accuracies by task-specific learning. However, estimation of large 3D deformations, for example present in inhale to exhale lung CT or interpatient abdominal MRI registration, is still a major challenge for the widely adopted U-Net-like network architectures. Even when using multi-level strategies, current state-of-the-art DL registration results do not yet reach the high accuracy of conventional frameworks. To overcome the problem of large deformations for deep learning approaches, in this work, we present GraphRegNet, a sparse keypoint-based geometric network for dense deformable medical image registration. Similar to the successful 2D optical flow estimation of FlowNet or PWC-Net we leverage discrete dense displacement maps to facilitate the registration process. In order to cope with enormously increasing memory requirements when working with displacement maps in 3D medical volumes and to obtain a well-regularised and accurate deformation field we 1) formulate the registration task as the prediction of displacement vectors on a sparse irregular grid of distinctive keypoints and 2) introduce our efficient GraphRegNet for displacement regularisation, a combination of convolutional and graph neural network layers in a unified architecture. In our experiments on exhale to inhale lung CT registration we demonstrate substantial improvements (TRE below 1.4 mm) over other deep learning methods. Our code is publicly available at https://github.com/multimodallearning/graphregnet.

Deep learning for skeleton-based action recognition

In order to further improve the recognition rate of skeleton action recognition, and to break the limitation that most of the previous deep learning-based methods have the input content of human joint coordinates, we propose a human skeleton action recognition algorithm combining geometric features and LSTM networks. In this paper, we propose a human skeleton action recognition algorithm that combines geometric features with LSTM network. The algorithm selects geometric features based on the distances between joints and selected lines as the input of the network, and introduces a time-selective LSTM network for training. The ability to select the most recognizable temporal features using the time-selective LSTM network is demonstrated in the SBU Interaction dataset and UT Kinect dataset, achieving 99.46% and 99.30% recognition rates, respectively. The experimental results demonstrate the effectiveness of the method for human skeleton-based action recognition.

Community lockdowns in social networks hardly mitigate epidemic spreading

Community lockdowns and travel restrictions are commonly employed to decelerate epidemic spreading. We here use a stochastic susceptible-infectious-recovered model on different social networks to determine when and to what degree such lockdowns are likely to be effective. Our research shows that community lockdowns are effective only if the links outside of the communities are virtually completely sealed off. The benefits of targeting specifically these links, as opposed to links uniformly at random across the whole network, are inferable only beyond 90% lockdown effectiveness. And even then the peak of the infected curve decreases by only 20% and its onset is delayed by a factor of 1.5. This holds for static and temporal social networks, regardless of their size and structural particularities. Networks derived from cell phone location data and online location-based social platforms yield the same results as a large family of hyperbolic geometric network models where characteristic path lengths, clustering, and community structure can be arbitrarily adjusted. The complex connectedness of modern human societies, which enables the ease of global communication and the lightning speeds at which news and information spread, thus makes it very difficult to halt epidemic spreading with top-down measures. We therefore emphasize the outstanding importance of endogenous self-isolation and social distancing for successfully arresting epidemic spreading.

Protein interaction interface region prediction by geometric deep learning.

Protein-protein interactions drive wide-ranging molecular processes, and characterizing at the atomic level how proteins interact (beyond just the fact that they interact) can provide key insights into understanding and controlling this machinery. Unfortunately, experimental determination of three-dimensional protein complex structures remains difficult and does not scale to the increasingly large sets of proteins whose interactions are of interest. Computational methods are thus required to meet the demands of large-scale, high-throughput prediction of how proteins interact, but unfortunately, both physical modeling and machine learning methods suffer from poor precision and/or recall. In order to improve performance in predicting protein interaction interfaces, we leverage the best properties of both data- and physics-driven methods to develop a unified Geometric Deep Neural Network, 'PInet' (Protein Interface Network). PInet consumes pairs of point clouds encoding the structures of two partner proteins, in order to predict their structural regions mediating interaction. To make such predictions, PInet learns and utilizes models capturing both geometrical and physicochemical molecular surface complementarity. In application to a set of benchmarks, PInet simultaneously predicts the interface regions on both interacting proteins, achieving performance equivalent to or even much better than the state-of-the-art predictor for each dataset. Furthermore, since PInet is based on joint segmentation of a representation of a protein surfaces, its predictions are meaningful in terms of the underlying physical complementarity driving molecular recognition. PInet scripts and models are available at https://github.com/FTD007/PInet. Supplementary data are available at Bioinformatics online.

Multi-derivative physical and geometric convolutional embedding networks for skeleton-based action recognition

Action involves rich geometric and physical properties hidden in the spatial structure and temporal dynamics. However, there is a lack of synergy in investigating these properties and their joint embedding in the existing literature. In this paper, we propose a multi-derivative physical and geometric embedding network (PGEN) for action recognition from skeleton data. We model the skeleton joint and edge information using multi-derivative physical and geometric features. Then, a physical and geometric embedding network is proposed to learn co-occurrence features from joints and edges, respectively, and construct a unified convolutional embedding space, where the physical and geometric properties can be integrated effectively. Furthermore, we adopt a multi-task learning framework to explore the inter-dependencies between the physical and geometric properties of the action, which significantly improves the discrimination of the learned features. The experiments on the NTU RGB+D, NTU RGB+D 120, and SBU datasets demonstrate the effectiveness of our proposed representation and modeling method.

A Preliminary Study of Applying Geometric Deep Learning in Brain Morphometry for Diagnosis of Alzheimer's Disease

A predictive model of Alzheimer's disease (AD) was established based on brain surface meshes and geometric deep learning, and its performance was evaluated. Seventy-six clinically diagnosed AD patients and 83 healthy older adults were enrolled and randomly assigned to the training set and the test set according to a 4-to-1 ratio. Brain surface mesh was constructed from 3-D T1-weighted high-resolution structural MR volumes of each participant. After applying a series of simplification to the surface meshes, the training set was fed into the geometric deep neural network for training. The performance of the prediction model was evaluated with the test set, and the evaluation metrics included accuracy, sensitivity and specificity. The prediction model trained on the right brain surface meshes with 6 000 faces achieved the best performance, with accuracy reaching 93.8%, sensitivity, 91.7%, and specificity, 94.1%. The evolution of the brain surface meshes during convolution and pooling revealed that AD patients had diffuse brain tissue loss compared with healthy older adults. Morphological brain analysis based on mesh data and geometric deep learning has great potential in the differential diagnosis of AD.

Connectivity and routes in large geometric network hypergraphs

Random geometric hypergraphs are considered as mathematical models of large wireless computer networks. The dependences of the mathematical expectation of the number of hyper-edges in random geometric hypergraphs on the radii of reliable reception / transmission of radio signals by network nodes, as well as on the number of vertices in the hyper- graph are studied. The concepts of the shortest route in a geometric hypergraph are discussed. Calculations of the probabil- ity of connectivity of large random geometric hypergraphs, the mathematical expectation of the diameter of hypergraphs and its change with a change in the radii of the nodes are carried out. The presentation of the results is accompanied by graphs.

MW-GAN: Multi-Warping GAN for Caricature Generation With Multi-Style Geometric Exaggeration.

Given an input face photo, the goal of caricature generation is to produce stylized, exaggerated caricatures that share the same identity as the photo. It requires simultaneous style transfer and shape exaggeration with rich diversity, and meanwhile preserving the identity of the input. To address this challenging problem, we propose a novel framework called Multi-Warping GAN (MW-GAN), including a style network and a geometric network that are designed to conduct style transfer and geometric exaggeration respectively. We bridge the gap between the style/landmark space and their corresponding latent code spaces by a dual way design, so as to generate caricatures with arbitrary styles and geometric exaggeration, which can be specified either through random sampling of latent code or from a given caricature sample. Besides, we apply identity preserving loss to both image space and landmark space, leading to a great improvement in quality of generated caricatures. Experiments show that caricatures generated by MW-GAN have better quality than existing methods.

Shear stress dependence of force networks in 3D dense suspensions.

The geometric organization and force networks of 3D dense suspensions that exhibit both shear thinning and thickening have been examined as a function of varying strength of interparticle attractive interactions using lubrication flow discrete element simulations. Significant rearrangement of the geometric topology does not occur at either the local or global scale as these systems transition across the shear thinning and shear thickening regimes. In contrast, massive rearrangements in the balance of attractive, lubrication, and contact forces are observed with interesting behavior of network growth and competition. In agreement with prior work, in shear thinning regions the attractive force is dominant, however as the shear thickening region is approached there is growth of lubrication forces. Lubrication forces oppose the attraction forces, but as viscosity continues to increase under increasing shear stress, the lubrication forces are dominated by contact forces that also resist attraction. Contact forces are the dominant interactions during shear thickening and are an order of magnitude higher than their values in the shear-thinning regime. At high attractive interaction strength, contact networks can form even under shear thinning conditions, however high shear stress is still required before contact networks become the driving mechanism of shear thickening. Analysis of the contact force network during shear thickening generally indicates a uniformly spreading network that rapidly forms across empty domains; however the growth patterns exhibit structure that is significantly dependent upon the strength of interparticle interactions, indicating subtle variations in the mechanism of shear thickening.

Geometric property-based convolutional neural network for indoor object detection

Indoor object detection is a very demanding and important task for robot applications. Object knowledge, such as two-dimensional (2D) shape and depth information, may be helpful for detection. In this article, we focus on region-based convolutional neural network (CNN) detector and propose a geometric property-based Faster R-CNN method (GP-Faster) for indoor object detection. GP-Faster incorporates geometric property in Faster R-CNN to improve the detection performance. In detail, we first use mesh grids that are the intersections of direct and inverse proportion functions to generate appropriate anchors for indoor objects. After the anchors are regressed to the regions of interest produced by a region proposal network (RPN-RoIs), we then use 2D geometric constraints to refine the RPN-RoIs, in which the 2D constraint of every classification is a convex hull region enclosing the width and height coordinates of the ground-truth boxes on the training set. Comparison experiments are implemented on two indoor datasets SUN2012 and NYUv2. Since the depth information is available in NYUv2, we involve depth constraints in GP-Faster and propose 3D geometric property-based Faster R-CNN (DGP-Faster) on NYUv2. The experimental results show that both GP-Faster and DGP-Faster increase the performance of the mean average precision.

Deep Neural Network-based Fusion and Natural Language Processing in Additive Manufacturing for Customer Satisfaction

Modern Machine learning fusion approaches tend to extract features depending on two techniques (hand-crafted feature and representation learning). Hand-crafted features can waste time and are not sufficient for downstream tasks. Unlike representation learning, we automatically learn features with minimum time and effort and are suitable for downstream tasks. In our paper, we provide work on graph neural network methods with details on classical graph embedding approaches and the different methods in neural graph networks such as graph filtering, graph pooling, and the learning parameter for graph following each technique with a general framework or mathematical proof for customer satisfaction. To satisfy customer's feel, this research employs NLP techniques. We describe the adversarial attacks and defenses on graph representation approaches. Also, advanced application of neural graph networks is reviewed, such as combinational optimization, learning program representation, physical system modeling, and natural language processing. Finally, the challenges in geometric neural networks and future research work have been introduced.