Global Structural Features Research Articles

Generating Structurally Complete Stylish Chinese Font Based on Semi-Supervised Model

Stylish fonts are widely utilized in social networks, driving research interest in stylish Chinese font generation. Most existing methods for generating Chinese characters rely on GAN-based deep learning approaches. However, persistent issues of mode collapse and missing structure in CycleGAN-generated results make font generation a formidable task. This paper introduces a unique semi-supervised model specifically designed for generating stylish Chinese fonts. By incorporating a small amount of paired data and stroke encoding as information into CycleGAN, our model effectively captures both the global structural features and local characteristics of Chinese characters. Additionally, an attention module has been added and refined to improve the connection between strokes and character structures. To enhance the model’s understanding of the global structure of Chinese characters and stroke encoding reconstruction, two loss functions are included. The integration of these components successfully alleviates mode collapse and structural errors. To assess the effectiveness of our approach, comprehensive visual and quantitative assessments are conducted, comparing our method with benchmark approaches on six diverse datasets. The results clearly demonstrate the superior performance of our method in generating stylish Chinese fonts.

Deep attention sampling hashing for efficient image retrieval

Hashing has received broad attention in large-scale image retrieval due to its appealing efficiency in computation and storage. Particularly, with the drawn of deep learning, much efforts have been directed towards using deep neural networks to learn feature representations and hash codes simultaneously, and the developed deep hashing methods have shown superior performance over conventional hashing methods. In this paper, we propose Deep Attention Sampling Hashing (DASH), a novel deep hashing method that yields high-quality hash codes to enable efficient image retrieval. Specifically, we employ two sub-networks in DASH, i.e., a master branch and a part branch, to capture global structure features and discriminative feature representations, respectively. Furthermore, we develop an Attention Sampler Module (ASM), which consists of an Object Region Extraction (ORE) block and an Informative Patch Generation (IPG) block, to yield richer informative image patches. The ORE block provides a well-designed multi-scale attentional fusion mechanism to highlight and extract the significant regions of images, and the IPG block employs a direction-specific shift mechanism to generate desired image patches with discriminative details. Both blocks could be seamlessly integrated into various convolutional neural network (CNN) architectures. Subsequently, we conduct knowledge distillation optimization to transfer the details learned by the part branch into the master branch to guide hash code learning. In addition, we design a Weibull quantization loss to minimize the information loss caused by binary quantization. The experimental results on three benchmark datasets demonstrate the effectiveness of the proposed DASH with respect to different evaluation metrics.

Structure Learning Via Meta-Hyperedge for Dynamic Rumor Detection

Online social networks have greatly facilitated our lives but have also propagated the spreading of rumours. Traditional works mostly find rumors from content, but content can be strategically manipulated to evade such detection, making these methods brittle. To improve the accuracy and robustness of rumor detection, we propose to integrate and exploit the content, propagation structure, and temporal relations because information in the networks always spreads dynamically with significant structures. In this paper, we propose a novel rumor detection framework in online temporal networks via structure learning. Specifically, to exploit the propagation structure, we propose a novel hyperedge walking strategy on a meta-hyperedge graph to learn the representations of sub-structures in the networks. Then a hyperedge expansion method is proposed to generate more global structural features. The expanded hyperedges are more hierarchical, making the learned structural embeddings more expressive. To make full use of content, we design a hypergraph learning model using hyperedge expansion to fuse node content with structural features and generate comprehensive representations for the entire graph. To exploit temporal relations, we design a masked temporal attention unit for learning the evolving patterns of the network. Extensive evaluations with six state-of-the-art baselines on two real-world datasets demonstrate the superiority of our solution.

Effect of hidden geometry and higher-order interactions on the synchronization and hysteresis behavior of phase oscillators on 5-clique simplicial assemblies.

The hidden geometry of simplicial complexes can influence the collective dynamics of nodes in different ways depending on the simplex-based interactions of various orders and competition between local and global structural features. We study a system of phase oscillators attached to nodes of four-dimensional simplicial complexes and interacting via positive/negative edges-based pairwise K_{1} and triangle-based triple K_{2}≥0 couplings. Three prototypal simplicial complexes are grown by aggregation of 5-cliques, controlled by the chemical affinity parameter ν, resulting in sparse, mixed, and compact architecture, all of which have 1-hyperbolic graphs but different spectral dimensions. By changing the interaction strength K_{1}∈[-4,2] along the forward and backward sweeps, we numerically determine individual phases of each oscillator and a global order parameter to measure the level of synchronization. Our results reveal how different architectures of simplicial complexes, in conjunction with the interactions and internal-frequency distributions, impact the shape of the hysteresis loop and lead to patterns of locally synchronized groups that hinder global network synchronization. Remarkably, these groups are differently affected by the size of the shared faces between neighboring 5-cliques and the presence of higher-order interactions. At K_{1}<0, partial synchronization is much higher in the compact community than in the assemblies of cliques sharing single nodes, at least occasionally. These structures also partially desynchronize at a lower triangle-based coupling K_{2} than the compact assembly. Broadening of the internal frequency distribution gradually reduces the synchronization level in the mixed and sparse communities, even at positive pairwise couplings. The order-parameter fluctuations in these partially synchronized states are quasicyclical with higher harmonics, described by multifractal analysis and broad singularity spectra.

GWBNER: A named entity recognition method based on character glyph and word boundary features for Chinese EHRs

Electronic Health Records (EHRs) contain unprecedented volumes of data related to health such as diagnosis and treatment information. Based on the technologies of named entity recognition (NER), EHRs mining has become a research focus in health domain. Nevertheless, the complex medical entities are challenging to be recognized, especially in Chinese EHRs. In this paper, a Glyph and Word Boundary-based Named Entity Recognition (GWBNER) method is presented, which takes into account both the Chinese character glyph and word boundary features in Chinese EHRs. Specifically, the character glyphs are utilized to capture the character-level global structural features, and the word boundaries are adopted to extract the word-level local structural features in Chinese. Therefore, the medical entities are fully recognized with rich semantic features from diverse perspectives in Chinese EHR texts. Finally, we conduct extensive experiments to evaluate the novel approach, and the F1 scores of 0.879, 0.998 and 0.962 are achieved on three EHR datasets, respectively. The experimental results demonstrate the optimal performance of the GWBNER method compared with the state-of-the-art models in NER community.

A Knowledge-based and Data-driven Approach for Predicting Acute Kidney Injury in Patients with Heart Failure.

It has great potential to integrate medical knowledge and electronic health record data for diagnosis prediction. However, present studies only utilized information from knowledge graphs, omitting potentially significant global graph structural features. In this study, we proposed a knowledge and data integrating modeling approach to reconstruct patient electronic health record data with graph structure and use medical knowledge as internal information of patient data to build a risk prediction model for acute kidney injury in patients with heart failure based on graph neural networks. Experimental results based on the MIMIC III data showed that the method proposed was superior to other baseline models in predicting the risk of acute kidney injury in heart failure patients, with an accuracy of 0.725 and an F1 score of 0.755. This study provides a novel approach to the disease risk prediction models that integrates medical knowledge and data.

HackGAN: Harmonious Cross-Network Mapping Using CycleGAN With Wasserstein–Procrustes Learning for Unsupervised Network Alignment

Network alignment (NA) that identifies equivalent nodes across networks is an effective tool for integrating knowledge from multiple networks. The state-of-the-art NA methods learn inter-network node similarities based on labeled anchor links, which are costly, time-consuming, and difficult to acquire. Therefore, a few unsupervised network alignment (UNA) methods propose solving NA problems without anchor links. However, most existing UNA methods rely on discriminative attributes to capture nodes’ similarities and are hard to obtain optimal one-to-one alignments. Toward these issues, this article proposes a novel method named HackGAN to solve the UNA problem solely based on the structural information. Specifically, HackGAN represents nodes with embeddings based on an unsupervised graph neural network (GNN) to capture their global and local structural features. After that, it initializes mapping functions to transform the embedding spaces of different networks into the same vector space by iteratively solving the Wasserstein–Procrustes problem. The mapping functions are then refined by an adversarial model with cycle-consistency and Sinkhorn distance losses to obtain optimized one-to-one mappings. Based on the distances between mapped embeddings, accurate and robust results are obtained with a collective alignment algorithm. Experimental comparisons on both synthetic and real-world datasets demonstrate the superiority of HackGAN.

Global population structure, genomic diversity and carbohydrate fermentation characteristics of clonal complex 119 (CC119), an understudied Shiga toxin-producing E. coli (STEC) lineage including O165:H25 and O172:H25.

Among Shiga toxin (Stx)-producing Escherichia coli (STEC) strains of various serotypes, O157:H7 and five major non-O157 STEC (O26:H11, O111:H8, O103:H2, O121:H19 and O145:H28) can be selectively isolated by using tellurite-containing media. While human infections by O165:H25 STEC strains have been reported worldwide, their detection and isolation are not easy, as they are not resistant to tellurite. Systematic whole-genome sequencing (WGS) analyses have not yet been conducted. Here, we defined O165:H25 strains and their close relatives, including O172:H25 strains, as clonal complex 119 (CC119) and performed a global WGS analysis of the major lineage of CC119, called CC119 sensu stricto (CC119ss), by using 202 CC119ss strains, including 90 strains sequenced in this study. Detailed comparisons of 13 closed genomes, including 7 obtained in this study, and systematic analyses of Stx phage genomes in 50 strains covering the entire CC119ss lineage, were also conducted. These analyses revealed that the Stx2a phage, the locus of enterocyte effacement (LEE) encoding a type III secretion system (T3SS), many prophages encoding T3SS effectors, and the virulence plasmid were acquired by the common ancestor of CC119ss and have been stably maintained in this lineage, while unusual exchanges of Stx1a and Stx2c phages were found at a single integration site. Although the genome sequences of Stx2a phages were highly conserved, CC119ss strains exhibited notable variation in Stx2 production levels. Further analyses revealed the lack of SpLE1-like elements carrying the tellurite resistance genes in CC119ss and defects in rhamnose, sucrose, salicin and dulcitol fermentation. The genetic backgrounds underlying these defects were also clarified.

Local image descriptor developed from Fechner’s law

Human perception of visual stimulus and its physical characteristics have a nonlinear-logarithmic relation as stated in Fechner’s law, which is a psychophysical law of perception (i.e., subjective sensation). Inspired by Fechner’s law, we use the salient differences among pixels in images as the physical characteristics to mimic the human pattern perception and propose a Fechner local image descriptor (FLID) for face image representation. FLID is a nonlinear descriptor, so the represented image is more hierarchical than the general linear one. In addition, a threshold method is used to divide the features into four different intervals, which can effectively reduce the effects of noise and illumination. However, considering that FLID only extracts the local features of single-scale blocks, we incorporate multi-scale block into FLID and propose a multi-scale block Fechner local image descriptor (MB-FLID) to extract more global structural features. In the experimental part, we verified the performance of FLID and MB-FLID on the Face Recognition Technology (FERET), extended Yale-B, Olivetti Research Laboratories, Yale, and AR face databases, and compared them with a host of other local feature descriptors. The experimental results show that the proposed FLID and MB-FLID outperform the compared descriptors.

Non-Rigid Point Cloud Matching Based on Invariant Structure for Face Deformation

In this paper, we present a non-rigid point cloud matching method based on an invariant structure for face deformation. Our work is guided by the realistic needs of 3D face reconstruction and re-topology, which critically need support for calculating the correspondence between deformable models. Our paper makes three main contributions: First, we propose an approach to normalize the global structure features of expressive faces using texture space properties, which decreases the variation magnitude of facial landmarks. Second, we make a modification to the traditional shape context descriptor to solve the problem of regional cross-mismatch. Third, we collect a dataset with various expressions. Ablation studies and comparative experiments were conducted to investigate the performance of the above work. In face deformable cases, our method achieved 99.89% accuracy on our homemade face dataset, showing superior performance over some other popular algorithms. In this way, it can help modelers to build digital humans more easily based on the estimated correspondence of facial landmarks, saving a lot of manpower and time.

Toward A No-reference Omnidirectional Image Quality Evaluation by Using Multi-perceptual Features

Compared to ordinary images, omnidirectional image (OI) usually has a broader view and a higher resolution, and image quality assessment (IQA) can help people to understand and improve their visual experience. However, the current IQA works cannot achieve good performance. To address this, we proposed a novel visual perception-based no-reference/blind omnidirectional image quality assessment (NR/B-OIQA) model. The gradient-based global structural features and gray-level co-occurrence matrix-based local structural features are combined together to highlight the rich quality-aware structural information. And a novel steganalysis real model-based color descriptor is extracted to reflect the color information that ignored in most IQA models. With a multi-scale visual perception, we take image entropy and the natural scene statistics features to convey the high-level semantics and quantify the unnaturalness of omnidirectional images. Finally, we apply support vector regression to predict the objective quality value based on the subjective scores and extracted all features. Experiments are conducted on OIQA and CVIQD2018 Databases, and the results illustrate that our model has more reliable performance and stronger competitiveness and receives better conformity with the subjective values.

DFSP: A fast and automatic distance field-based stem-leaf segmentation pipeline for point cloud of maize shoot.

The 3D point cloud data are used to analyze plant morphological structure. Organ segmentation of a single plant can be directly used to determine the accuracy and reliability of organ-level phenotypic estimation in a point-cloud study. However, it is difficult to achieve a high-precision, automatic, and fast plant point cloud segmentation. Besides, a few methods can easily integrate the global structural features and local morphological features of point clouds relatively at a reduced cost. In this paper, a distance field-based segmentation pipeline (DFSP) which could code the global spatial structure and local connection of a plant was developed to realize rapid organ location and segmentation. The terminal point clouds of different plant organs were first extracted via DFSP during the stem-leaf segmentation, followed by the identification of the low-end point cloud of maize stem based on the local geometric features. The regional growth was then combined to obtain a stem point cloud. Finally, the instance segmentation of the leaf point cloud was realized using DFSP. The segmentation method was tested on 420 maize and compared with the manually obtained ground truth. Notably, DFSP had an average processing time of 1.52 s for about 15,000 points of maize plant data. The mean precision, recall, and micro F1 score of the DFSP segmentation algorithm were 0.905, 0.899, and 0.902, respectively. These findings suggest that DFSP can accurately, rapidly, and automatically achieve maize stem-leaf segmentation tasks and could be effective in maize phenotype research. The source code can be found at https://github.com/syau-miao/DFSP.git.

Transformer-based comparative multi-view illegal transaction detection.

In recent years, as the Ether platform has grown by leaps and bounds. Numerous unscrupulous individuals have used illegal transaction to defraud large sums of money, causing billions of dollars of losses to investors worldwide. Facing the endless stream of the illegal transaction based on Ether smart contracts problems, such as illegal transaction, money laundering, financial fraud, phishing. Currently, illegal transaction are only detected by a single view of the smart contract's contract code view feature and account transaction view feature, which is not only incomplete, but also not fully representative of the smart contract's features. More importantly, the single view detection model cannot accurately capture the global structure and semantic features between the Tokens of the view features. In this case, it is particularly important that all view features are shared among themselves. In this paper, we investigate a Transformer-based model for contrasting illegal transaction detection networks under multiple views (TranMulti-View Net). The model in this paper is based on Transformer to learn a multi-view fusion representation, which aims to maximise the fusion of the interaction information of different view features under the same condition. In this model we first use the Transformer model to learn global structure and semantic features from a sequence of Tokens tokenised by a view, capturing the remote dependencies of Tokens in the view features, and then we share the contract code view features and the account transaction view features across all views to learn important semantic information between views from each other. In addition, we find that the approach of semi-supervised training of multi-view features using contrast learning outperforms the scheme of prediction based on direct fusion of different view features, resulting in stronger correlation between view features. As a result, the underlying semantic information can be captured more accurately, leading to more accurate predictions of illegal transaction. The experimental results show that our proposed TranMulti-View Net obtains good detection results with a Precision score of 98%.

Global and Local Structure Network for Image Classification

Principal Component Analysis Network (PCANet) is a feature learning algorithm which is widely used in face recognition and object classification. However, original PCANet still has some shortages. One is that PCA algorithm only extracts features by considering the global structure. The other lies in that the original PCANet only employs one particular single layer convolutional results, which loses the information of other convolutional layers. In this paper, we propose a new simple and efficient convolutional neural network called global and local structure network (GLSNet) to address the problems. The network extracts the features both from the global structure and the local structure of the original data space. Specifically, a principal component analysis (PCA) convolutional layer which learns the filters by PCA algorithm is used to remove the noises and redundant information at the first stage. Then at the second stage, another PCA convolution is added to extract features by considering the global structure. As for the local structure, we use the neighborhood preserving embedding (NPE) algorithm to learn the convolutional filters. At the output stage, the global structure feature extracted by PCA convolution and the local structure feature extracted by NPE convolution is concatenated as a united feature. Furthermore, the first layer convolutional feature is also taken into consideration to obtain shallow-level information. Finally, these features are concatenated as a united feature, and a spatial pyramid pooling layer is followed to pool above the united features. To test the effectiveness of the proposed algorithm, the experiments on some image datasets, including three types: human face dataset, object dataset, and handprinted dataset, proceeded. And it performs better than the original PCANet and some improvement algorithms of PCANet, such as PLDANet, and MMPCANet.

Graph-Based Structural Deep Spectral-Spatial Clustering for Hyperspectral Image

As a fundamental task in hyperspectral image (HSI) applications, HSI clustering has been extensively researched over the years. Inspired by the capability of graphs in structure modeling, some recent works have exploited the local graph-structured information among pixels to boost the HSI clustering performance. However, these works only explore shallow graph structures for representation learning, while ignoring the beneficial high-order as well as global structure characteristics of HSI pixels, thus resulting in limited clustering performance. In this paper, a novel graph-based structural deep spectral-spatial clustering network (GSDSSC) is proposed to sufficiently explore the structure information among pixels. Specifically, a self-expression embedded multi-graph auto-encoder is proposed to explore high-order structure associations among pixels, thereby capturing robust spectral-spatial features and global clustering structure. Besides, to learn more compact spectral-spatial features for clustering, a global structure-guided optimization mechanism is further designed to constrain the local feature distribution with global clustering structure. Extensive experiments on three public HSI benchmark datasets illustrate the superior performance of the proposed GSDSSC over state-of-the-art HSI clustering methods.

Elucidation of pH-Induced Protein Structural Changes: A Combined 2D IR and Computational Approach.

The acid-base behavior of amino acids plays critical roles in several biochemical processes. Depending on the interactions with the protein environment, the pKa values of these amino acids shift from their respective solution values. As the side chains interact with the polypeptide backbone, a pH-induced change in the protonation state of aspartic and glutamic acids might significantly influence the structure and stability of a protein. In this work, we have combined two-dimensional infrared spectroscopy and molecular dynamics simulations to elucidate the pH-induced structural changes in an antimicrobial enzyme, lysozyme, over a wide range of pH. Simultaneous measurements of the carbonyl signals arising from the backbone and the acidic side chains provide detailed information about the pH dependence of the local and global structural features. An excellent agreement between the experimental and the computational results allowed us to obtain a residue-specific molecular understanding. Although lysozyme retains the helical structure for the entire pH range, one distinct loop region (residues 65-75) undergoes local structural deformation at low pH. Interestingly, combining our experiments and simulations, we have identified the aspartic acid residues in lysozyme, which are influenced the most/least by pH modulation.

BepiPred-3.0: Improved B-cell epitope prediction using protein language models.

B-cell epitope prediction tools are of great medical and commercial interest due to their practical applications in vaccine development and disease diagnostics. The introduction of protein language models (LMs), trained on unprecedented large datasets of protein sequences and structures, tap into a powerful numeric representation that can be exploited to accurately predict local and global protein structural features from amino acid sequences only. In this paper, we present BepiPred-3.0, a sequence-based epitope prediction tool that, by exploiting LM embeddings, greatly improves the prediction accuracy for both linear and conformational epitope prediction on several independent test sets. Furthermore, by carefully selecting additional input variables and epitope residue annotation strategy, performance was further improved, thus achieving unprecedented predictive power. Our tool can predict epitopes across hundreds of sequences in minutes. It is freely available as a web server and a standalone package at https://services.healthtech.dtu.dk/service.php?BepiPred-3.0 with a user-friendly interface to navigate the results.

MDG: Fusion learning of the maximal diffusion, deep propagation and global structure features of fake news

Fake news detection has attracted much attention in the past five years. Some early studies use news content to detect fake news. As content feature-based methods lack generalizability for fake news detection, researchers explored methods that use social context features of news. However, existing studies only use single-feature (either propagation or structure) detection, but none uses diffusion group information (e.g., the group with the most user responses in the news events) in detecting fake news. Thus, this research develops a fake news detection framework named MDG, which learns and integrates the maximal diffusion feature with the deep propagation and global structure features. To validate the performance of our framework, two sets of experiments are designed on three publicly available datasets, i.e., Twitter15, Twitter16, and Weibo. Experimental results show that our framework outperforms the state-of-the-art methods in accuracy and F1-macro performance. Moreover, our framework is especially suitable for multi-lingual fake news detection.

Cryo-EM structures of cancer-specific helical and kinase domain mutations of PI3Kα

Phosphoinositide 3-kinases (PI3Ks) are a family of lipid kinases that perform multiple and important cellular functions. The protein investigated here belongs to class IA of the PI3Ks; it is a dimer consisting of a catalytic subunit, p110α, and a regulatory subunit, p85α, and is referred to as PI3Kα. The catalytic subunit p110α is frequently mutated in cancer. The mutations induce a gain of function and constitute a driving force in cancer development. About 80% of these mutations lead to single-amino-acid substitutions in one of three sites of p110α: two in the helical domain of the protein (E542K and E545K) and one at the C-terminus of the kinase domain (H1047R). Here, we report the cryo-electron microscopy structures of these mutants in complex with the p110α-specific inhibitor BYL-719. The H1047R mutant rotates its sidechain to a new position and weakens the kα11 activation loop interaction, thereby reducing the inhibitory effect of p85α on p110α. E542K and E545K completely abolish the tight interaction between the helical domain of p110α and the N-terminal SH2 domain of p85α and lead to the disruption of all p85α binding and a dramatic increase in flexibility of the adaptor-binding domain (ABD) in p110α. Yet, the dimerization of PI3Kα is preserved through the ABD-p85α interaction. The local and global structural features induced by these mutations provide molecular insights into the activation of PI3Kα, deepen our understanding of the oncogenic mechanism of this important signaling molecule, and may facilitate the development of mutant-specific inhibitors.

NeRD: a multichannel neural network to predict cellular response of drugs by integrating multidimensional data

BackgroundConsidering the heterogeneity of tumors, it is a key issue in precision medicine to predict the drug response of each individual. The accumulation of various types of drug informatics and multi-omics data facilitates the development of efficient models for drug response prediction. However, the selection of high-quality data sources and the design of suitable methods remain a challenge.MethodsIn this paper, we design NeRD, a multidimensional data integration model based on the PRISM drug response database, to predict the cellular response of drugs. Four feature extractors, including drug structure extractor (DSE), molecular fingerprint extractor (MFE), miRNA expression extractor (mEE), and copy number extractor (CNE), are designed for different types and dimensions of data. A fully connected network is used to fuse all features and make predictions.ResultsExperimental results demonstrate the effective integration of the global and local structural features of drugs, as well as the features of cell lines from different omics data. For all metrics tested on the PRISM database, NeRD surpassed previous approaches. We also verified that NeRD has strong reliability in the prediction results of new samples. Moreover, unlike other algorithms, when the amount of training data was reduced, NeRD maintained stable performance.ConclusionsNeRD’s feature fusion provides a new idea for drug response prediction, which is of great significance for precise cancer treatment.