Search

The Dunhuang ancient manuscripts are an excellent and precious cultural heritage of humanity. However, due to their age, the vast majority of these treasures are damaged and fragmented. Faced with a wide range of sources and numerous fragments, the process of restoration generally involves two core elements: sibling fragments identification and fragment assembly. Currently, fragment restoration still heavily relies on manual labor. During the long practice, a consensus has been reached on the importance of edge features for not only assembly but also for identification. However, accurate extraction of edge features and their use for efficient identification requires extensive knowledge and strong memory. This is a challenge for the human brain. So that in previous studies, fragment edge features have been used for assembly validation but rarely for identification. Therefore, an edge matcher is proposed, working like a bloodhound, capable of "sniffing out" specific "flavors" in edge features and performing efficient sibling fragment identification accordingly, providing guidance when experts perform entity assembly subsequently. Firstly, the fragmented images are standardized. Secondly, traditional methods are used to compress the representation of fragment edges and obtain paired local edge images. Finally, these images are fed into the edge matcher for classification discrimination, which is a CNN-based pairwise similarity metric model proposed in this paper, introducing residual blocks and depthwise separable convolutions, and adding multi-scale convolutional layers. With the edge matcher, a complex matching problem is successfully transformed into a simple classification problem. In the absence of a standard public dataset, a Dunhuang manuscript fragment edge dataset is constructed. Experiments are conducted on that dataset, and the accuracy, precision, recall, and F1 scores of the edge matcher all exceeded 97%. The effectiveness of the edge matcher is demonstrated by comparative experiments, and the rationality of the method design is verified by ablation experiments. The method combines traditional methods and deep learning methods to creatively use the edge geometric features of fragments for sibling fragment identification in a natural rather than coded way, making full use of the computer's computational and memory capabilities. The edge matcher can significantly reduce the time and scope of searching, matching, and inferring fragments, and assist in the reconstruction of Dunhuang ancient manuscript fragments. [ABSTRACT FROM AUTHOR]

Aiming at the problem of strong mining pressure in the near shallow buried and super-large mining height face, and considering the first 108 working faces in Jinjitan Coal Mine as the engineering background, the movement pattern and pressure distribution characteristics of the overlying rock layer on the 8.2 m fully mechanized mining face were analyzed from the perspective of theoretical analysis, field monitoring data, and numerical simulation. The results of the spatial structure mechanics model and FLAC3D numerical model of the mining face with a super-large mining height established in this study, which indicated that the mining operation of the mining face with a super-large mining height experienced rock dynamic load pressure and large-small periodic pressure phenomena. The fracture of the lower keystone beam leads to a small pressure cycle, and a large pressure cycle occurs when both the upper and lower keystone beams are fractured. Generally, the step distance during the size cycle is about twice the normal cycle. The site monitoring data shows that the initial incoming pressure step is 102 m and the periodic incoming pressure step is about 28.7 m, which is consistent with the theoretical value. When the working surface advances slowly, the dynamic load factor is smaller, and the incoming pressure step and duration are shorter, and vice versa. The research results are of important reference significance for mine pressure law and disaster prevention in similar conditions. [ABSTRACT FROM AUTHOR]

Purpose: Long-term hospitalized patients with schizophrenia (SCZ) are vulnerable to physical illness, leading to impaired life expectancy and treatment outcomes. There are few studies on the influence of non-alcoholic fatty liver disease (NAFLD) in long-term hospitalized patients. This study aimed to investigate the prevalence of and influence factors for NAFLD in hospitalized patients with SCZ. Patients and Methods: This cross-sectional retrospective study included 310 patients who had experienced long-term hospitalization for SCZ. NAFLD was diagnosed based on the results of abdominal ultrasonography. The T-test, Mann–Whitney U-test, correlation analysis, and logistic regression analysis were used to determine the influence factors for NAFLD. Results: Among the 310 patients who had experienced long-term hospitalization for SCZ, the prevalence of NAFLD was 54.84%. Antipsychotic polypharmacy (APP), body mass index (BMI), hypertension, diabetes, total cholesterol (TC), apolipoprotein B (ApoB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), triglycerides (TG), uric acid, blood glucose, gamma-glutamyl transpeptidase (GGT), high-density lipoprotein, neutrophil-to-lymphocyte ratio, and platelet-to-lymphocyte ratio significantly differed between the NAFLD and non-NAFLD groups (all P< 0.05). Hypertension, diabetes, APP, BMI, TG, TC, AST, ApoB, ALT, and GGT were positively correlated with NAFLD (all P< 0.05). The results of the logistic regression analysis indicated that APP, diabetes, BMI, ALT, and ApoB were the influence factors for NAFLD in patients with SCZ. Conclusion: Our results suggest a high prevalence of NAFLD among patients hospitalized long-term due to severe SCZ symptoms. Moreover, a history of diabetes, APP, overweight/obese status, and increased levels of ALT and ApoB were identified as negative factors for NAFLD in these patients. These findings may provide a theoretical basis for the prevention and treatment of NAFLD in patients with SCZ and contribute to the development of novel targeted treatments. [ABSTRACT FROM AUTHOR]

The relationship between cancer and iron metabolism has received more attention, but there is less research on nutritional support to improve the nutritional and inflammatory status of cancer patients by regulating iron metabolism. A survey was conducted involving 90 patients with lung cancer. The control group was given routine hospitalization diet and the oral nutrition supplement (ONS) group was given oral nutrient solution on the basis of routine hospitalization diet. It lasted for 2 weeks. The changes of iron metabolism, nutritional status, and inflammatory indicators of the two groups were compared. After 2 weeks of intervention, the levels of ALB and PA in the ONS group were increased than that in the control group (P <.05), the inflammation was significantly decreased in the ONS group (P <.05), and the levels of Hb, SI, and TIBC in the ONS group were higher than that in the control group, while the SF was significantly reduced (P <.05). ONS could improve iron metabolism, nutritional status, and inflammatory levels. We speculate that the decrease in inflammation may be due to the changes of iron metabolism, thereby improving nutritional status. It may become a new target for tumor treatment. [ABSTRACT FROM AUTHOR]

Graph-based clustering plays an important role in the clustering area. Recent studies about graph neural networks (GNN) have achieved impressive success on graph-type data. However, in general clustering tasks, the graph structure of data does not exist such that GNN can not be applied to clustering directly and the strategy to construct a graph is crucial for performance. Therefore, how to extend GNN into general clustering tasks is an attractive problem. In this paper, we propose a graph auto-encoder for general data clustering, AdaGAE, which constructs the graph adaptively according to the generative perspective of graphs. The adaptive process is designed to induce the model to exploit the high-level information behind data and utilize the non-euclidean structure sufficiently. Importantly, we find that the simple update of the graph will result in severe degeneration, which can be concluded as better reconstruction means worse update. We provide rigorous analysis theoretically and empirically. Then we further design a novel mechanism to avoid the collapse. Via extending the generative graph models to general type data, a graph auto-encoder with a novel decoder is devised and the weighted graphs can be also applied to GNN. AdaGAE performs well and stably in different scale and type datasets. Besides, it is insensitive to the initialization of parameters and requires no pretraining. [ABSTRACT FROM AUTHOR]

Graph-based clustering plays an important role in the clustering area. Recent studies about graph neural networks (GNN) have achieved impressive success on graph-type data. However, in general clustering tasks, the graph structure of data does not exist such that GNN can not be applied to clustering directly and the strategy to construct a graph is crucial for performance. Therefore, how to extend GNN into general clustering tasks is an attractive problem. In this paper, we propose a graph auto-encoder for general data clustering, AdaGAE, which constructs the graph adaptively according to the generative perspective of graphs. The adaptive process is designed to induce the model to exploit the high-level information behind data and utilize the non-euclidean structure sufficiently. Importantly, we find that the simple update of the graph will result in severe degeneration, which can be concluded as better reconstruction means worse update. We provide rigorous analysis theoretically and empirically. Then we further design a novel mechanism to avoid the collapse. Via extending the generative graph models to general type data, a graph auto-encoder with a novel decoder is devised and the weighted graphs can be also applied to GNN. AdaGAE performs well and stably in different scale and type datasets. Besides, it is insensitive to the initialization of parameters and requires no pretraining. [ABSTRACT FROM AUTHOR]

A large population and rapid urbanization dramatically promote the heating supply demand, the combined heating and power (CHP) system for energy cascade utilization came into being. However, the research on the recovery and utilization of condensing heat, the exploration of the coupling law between power generation and heating supply, and the influence of heat source parameters on thermo-economic performance are still insufficient. To this end, two combined heating and power (CHP) systems coupled with an organic Rankine cycle (ORC) and vapor compression cycle (VCC) are proposed, and their thermodynamic and economic performances are optimized and analyzed by the laws of thermodynamics. Results show that the increase of the volume flow will increase the power generation and heating supply quantity of the system, and there is an optimal evaporation temperature range of 130–140 °C to optimize the performance of the system. The increase of heat source temperature will improve the economic performance of the system, but it will reduce the exergetic efficiency. Therefore, two factors should be comprehensively considered in practical engineering. There is mutual exclusivity between the net power output of the system and the heating supply quantity, it should be reasonably allocated according to the actual needs of users in engineering applications. In addition, the exergetic efficiency of the two systems can reach more than 60%, and the energy utilization rate is high, which indicates that the cascade utilization mode is reasonable. [ABSTRACT FROM AUTHOR]

Surrounding rock control and support stability in the process of coal seam mining in ultra-large height mining face are the key to normal mine operation. In this study, the roof movement and deformation of an ultra-large height mining face are analyzed, and the working resistance of the ultra-large height mining face is obtained by introducing the equivalent immediate roof. By analyzing the coal wall spalling, the multiple positions of the spalling and the required support force of the support are obtained. At the same time, ultra-large height supports are more prone to instability problems. In this study, the stability of the ultra-large height supports was analyzed by establishing a mechanical model. The results show that: 1. The overturning limit angle of support has a hyperbolic relationship with the center of gravity. 2. Under the condition of ultra-large height, the increase in the base width of the bracket significantly improves the stability of the supports. 3. The sliding limit angle of support is positively correlated with the support load and the friction coefficient between the support and the floor. The above conclusions can provide guidance on the selection of supports and the adoption of measures to enhance the stability of the supports during use under ultra-large height conditions. The working resistance of the ultra-large height supports in the 108 mining face of the Jinjitan Coal Mine was monitored. The monitoring results show that: The average resistance of the supports is 22.6 MPa. The selected supports can meet the stability requirements of the working face support. The frequency of mining resistance in 0~5 MPa accounts for 28.38%, which indicates that some supports are insufficient for the initial support force during the moving process. Furthermore, the stability of the supports can be enhanced by adjusting the moving process. This study provides a reference for the selection of supports in ultra-large height mining faces and proposes measures to enhance the stability of the supports, which provides guidance for the safe mining of coal in ultra-large height mining faces. [ABSTRACT FROM AUTHOR]

Stego-images are often contaminated by interchannel noise or active noise attack when communicating on the Web. And it is challenging to restore embedded image from corrupted stego-image. This paper studies a kNN-bit approximation algorithm to remove noises in embedded image. The proposed algorithm distinguishes reliable bits from extracted bits, and estimates pixel values by keeping reliable bits unchanged and correcting unreliable bits. Specifically, the 8th (highest) unreliable bit of a pixel can be approximated with its nearest neighbor pixels. And then, if an unreliable bit locates at any one of the $5^{th}\sim 7^{th}$ 5 t h ∼ 7 t h bits of a pixel, it is adjusted with two nearest neighbors of the pixel, where the pixel is in-between these two nearest neighbors. Finally, for other unreliable bits, each one is approximated by the maximum and minimum possible values of nearest neighbors of its pixel. We conduct experiments for illustrating the efficiency, and demonstrate that the proposed algorithm can recover the embedded images with good visual quality from corrupted stego-images. [ABSTRACT FROM AUTHOR]

Hyperspectral unmixing is an essential research topic for spectral data analysis due to the existence of mixed pixels. Recently, many methods based on sparse nonnegative matrix factorization (NMF) have been widely used for unmixing by incorporating similarity preserving. However, most of them conduct the similarity learning and unmixing by using two separate steps, which may lead to the case that the learned similarity matrix is not the optimal one for unmixing. Thus, the performance of unmixing would be influenced and become undesirable. To alleviate this problem, we propose an adaptive relationship preserving-based sparse NMF (ARP-NMF) for hyperspectral unmixing. Typically, we regard similarity learning and unmixing as an alternative optimization process. During this process, the learned similarity weights and unmixing results can be mutually improved. Thus, our proposed method can learn the optimal similarity weights for unmixing and obtain better generalization ability for different hyperspectral images than traditional methods. Moreover, by using the spectral and spatial local structure, our ARP-NMF method effectively preserves structure consistency between pixels and abundances. Experimental results both on the synthetic data and the real data reveal that our proposed method outperforms several representative sparse NMF-based methods. [ABSTRACT FROM AUTHOR]

Multiview clustering partitions data into different groups according to their heterogeneous features. Most existing methods degenerate the applicability of models due to their intractable hyper-parameters triggered by various regularization terms. Moreover, traditional spectral based methods always encounter the expensive time overheads and fail in exploring the explicit clusters from graphs. In this paper, we present a scalable and parameter-free graph fusion framework for multiview clustering, seeking for a joint graph compatible across multiple views in a self-supervised weighting manner. Our formulation coalesces multiple view-wise graphs straightforward and learns the weights as well as the joint graph interactively, which could actively release the model from any weight-related hyper-parameters. Meanwhile, we manipulate the joint graph by a connectivity constraint such that the connected components indicate clusters directly. The designed algorithm is initialization-independent and time-economical which obtains the stable performance and scales well with the data size. Substantial experiments on toy data as well as real datasets are conducted that verify the superiority of the proposed method compared to the state-of-the-arts over the clustering performance and time expenditure. [ABSTRACT FROM AUTHOR]

The objective was to investigate the safety and efficacy of the nutrition protocol in enhanced recovery after surgery (ERAS) of hypopharyngeal cancer. Protocol focus was patient consumption of nutritional supplements perioperatively. In this retrospective study, a total of 78 patients with hypopharyngeal cancer were divided into the ERAS group (n = 39) and the control group (n = 39). The data were collected from two groups of three time points: 1 day before surgery, 1 day after surgery, and 7 days after surgery. The difference between two groups of the nutritional and immune status, postoperative exhaust time, hospitalization expense and hospitalization time were compared. The nutritional and immune status in the ERAS group were better than that in the control group at 7 days after surgery (P <.05); The hospitalization expense and hospitalization time in the ERAS group were lower comparing with the control group (P <.05). Our nutrition protocol is effective and safe in ERAS of patients with hypopharyngeal cancer. It's significant to implement ERAS of hypopharyngeal cancer patients with nutritional protocol during peroperative period, which will improve immune system, maintain health metabolic functions, and reduce the hospitalization time as well as the hospitalization expense. [ABSTRACT FROM AUTHOR]

Excessive alcohol and iron intake can reportedly cause liver damage. In the present study, we investigated the effect of Lactobacillus casei on liver injury in rats co‐exposed to alcohol and iron and evaluated its possible mechanism. Sixty male Wistar rats were randomly divided into three groups for 12 weeks: the Control group (administered normal saline by gavage and provided a normal diet); alcohol +iron group (Model group, treated with alcohol [3.5–5.3 g/kg/day] by gavage and dietary iron [1,500 mg/kg]); Model group supplemented with L. casei (8 × 108 CFU kg−1 day−1) (L. casei group). Using hematoxylin and eosin (HE) staining and transmission electron microscopy, we observed that L. casei supplementation could alleviate disorders associated with lipid metabolism, inflammation, and intestinal mucosal barrier injury. Moreover, levels of serum alanine aminotransferase, gamma‐glutamyl transferase, triglyceride (TG), and hepatic TG were significantly increased in the model group; however, these levels were significantly decreased following the 12‐week L. casei supplementation. In addition, we observed notable improvements in intestinal mucosal barrier function and alterations in T lymphocyte subsets and natural killer cells in L. casei‐treated rats when compared with the model group. Furthermore, L. casei intervention alleviated serum levels of tumor necrosis factor‐α and interleukin‐1β, accompanied by decreased serum endotoxin levels and downregulated expression of toll‐like receptor 4 and its related molecules MyD88, nuclear factor kappa‐B p65, and TNF‐α. Accordingly, supplementation with L. casei could effectively improve liver injury induced by the synergistic interaction between alcohol and iron. The underlying mechanism for this improvement may be related to immune regulation and inhibition of enterogenic endotoxin‐mediated inflammation. [ABSTRACT FROM AUTHOR]

Recently, remote sensing image captioning (RSIC) has drawn an increasing attention. In this field, the encoder–decoder-based methods have become the mainstream due to their excellent performance. In the encoder–decoder framework, the convolutional neural network (CNN) is used to encode a remote sensing image into a semantic feature vector, and a sequence model such as long short-term memory (LSTM) is subsequently adopted to generate a content-related caption based on the feature vector. During the traditional training stage, the probability of the target word at each time step is forcibly optimized to 1 by the cross entropy (CE) loss. However, because of the variability and ambiguity of possible image captions, the target word could be replaced by other words like its synonyms, and therefore, such an optimization strategy would result in the overfitting of the network. In this article, we explore the overfitting phenomenon in the RSIC caused by CE loss and correspondingly propose a new truncation cross entropy (TCE) loss, aiming to alleviate the overfitting problem. In order to verify the effectiveness of the proposed approach, extensive comparison experiments are performed on three public RSIC data sets, including UCM-captions, Sydney-captions, and RSICD. The state-of-the-art result of Sydney-captions and RSICD and the competitive results of UCM-captions achieved by TCE loss demonstrate that the proposed method is beneficial to RSIC. [ABSTRACT FROM AUTHOR]

The variation of the in situ stress state is closely related to various factors. In situ stress state is also an important indicator to guide mining production. The study of in situ stress measurement and its distribution characteristics has always been a basic and very important work in mine production. In this study, the deep mines of Linyi Mining Area were considered as the research object. In this regard, the stress distribution law of each mine was studied. We found that the relationship between principal stresses was σH > σ v > σh, which belongs to the strike-slip stress regime. In this stress regime, the lateral Earth pressure coefficient was greater than one, and the magnitude of the three principal stresses all showed an increasing trend with the increase of depth. The maximum horizontal stress direction of the Gucheng Coal Mine, Guotun Coal Mine, and Pengzhuang Coal Mine was NW-SE under the influence of regional geological structure, while the maximum horizontal stress direction of Wanglou Coal Mine was NE-SW under the influence of local geological structure. Besides, the relationship between mine in situ stress and mine geological structure, the impact of original rock stress on stope stability, and the effect of original rock stress on floor water inrushing were also investigated. We believe that the research results are beneficial to mine disaster prevention and safety production. [ABSTRACT FROM AUTHOR]

Using the self-developed three-axis servo fluid-solid coupling system with gas-solid coupling of gas-bearing coal, the variation law of the permeability of gas coal under the stress cycle loading and unloading path was studied. The qualitative and quantitative relationships between permeability, axial force, and radial stress of gas-bearing coals were established, and the variation law of permeability of gas-bearing coals was discussed. The results show that (1) different cyclic loading and unloading stress paths correspond to the permeability characteristics of different gas-bearing coals. (2) Permeability of gas-bearing coal decreases with the increase of axial stress and radial stress, and it has a logarithmic function with axial stress and radial stress. This shows that axial stress and radial stress are important factors affecting the permeability characteristics of gas-bearing coal. (3) Under the same stress loading and unloading conditions, the axial stress is less than radial stress on the permeability of gas-bearing coal. In the cyclic loading and unloading axial stress process, the permeability of the gas-bearing coal varies by a smaller extent than the cyclically unloaded confining force. (4) The cumulative damage rate of gas-bearing coal under axial stress gradually increases with the increase of the number of cycles of loading and unloading, and the rate of the cumulative damage rate of permeability is less than the corresponding rate of radial stress. [ABSTRACT FROM AUTHOR]

The fusion of hyperspectral (HS) and multispectral (MS) images designed to obtain high-resolution HS (HRHS) images is a very challenging work. A series of solutions has been proposed in recent years. However, the similarity in the structure of the HS image has not been fully used. In this article, we present a novel HS and MS image-fusion method based on nonlocal low-rank tensor approximation and sparse representation. Specifically, the HS image and the MS image are considered the spatially and spectrally degraded versions of the HRHS image, respectively. Then, the nonlocal low-rank constraint term is adopted in order to form the nonlocal similarity and the spatial–spectral correlation. Meanwhile, we add the sparse constraint term to describe the sparsity of abundance. Thus, the proposed fusion model is established and its optimization is solved by alternative direction method of multipliers (ADMM). The experimental results on three synthetic data sets and one real data set show the advantages of the proposed method over several state-of-the-art competitors. [ABSTRACT FROM AUTHOR]

Collective behavior, from murmurations to synchronized beating of heart cells, governs some of the most beautiful and important aspects of nature. Likewise, cooperation—the act of sacrificing personal benefits for the common good—is one of the pillars of social evolution, and it is the basis for the emergence of collective organized actions from single-cell organisms to modern human societies. Here we merge these two phenomena into a single model, considering an ensemble of networked oscillators, where each oscillator can be either a cooperator or a defector, and with only cooperators contributing to synchrony. At the same time, the value of the order parameter in the neighborhood of each oscillator is considered as an effective local temperature which determines the strategy updating procedure in the evolutionary game. The emergence of cooperation is thus intertwined with that of synchronization, producing a novel and fascinating dynamics which includes a double explosive transition. [ABSTRACT FROM AUTHOR]

Existing video summarization approaches mainly concentrate on the sequential or structural characteristic of video data. However, they do not pay enough attention to the video summarization task itself. In this article, we propose a meta learning method for performing task-driven video summarization, denoted by MetaL-TDVS, to explicitly explore the video summarization mechanism among summarizing processes on different videos. Particularly, MetaL-TDVS aims to excavate the latent mechanism for summarizing video by reformulating video summarization as a meta learning problem and promote the generalization ability of the trained model. MetaL-TDVS regards summarizing each video as a single task to make better use of the experience and knowledge learned from processes of summarizing other videos to summarize new ones. Furthermore, MetaL-TDVS updates models via a twofold backpropagation, which forces the model optimized on one video to obtain high accuracy on another video in every training step. Extensive experiments on benchmark datasets demonstrate the superiority and better generalization ability of MetaL-TDVS against several state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Convolutional Neural Network (CNN) has shown their advantages in salient object detection. CNN can generate great saliency maps because it can obtain high-level semantic information. And the semantic information is usually achieved by stacking multiple convolutional layers and pooling layers. However, multiple pooling operations will reduce the size of the feature map and easily blur the boundary of the salient object. Therefore, such operations are not beneficial to generate great saliency results. To alleviate this issue, we propose a novel edge information-guided hierarchical feature fusion network (HFFNet). Our network fuses features hierarchically and retains accurate semantic information and clear edge information effectively. Specifically, we extract image features from different levels of VGG. Then, we fuse the features hierarchically to generate high-level semantic information and low-level edge information. In order to retain better information at different levels, we adopt a one-to-one hierarchical supervision strategy to supervise the generation of low-level information and high-level information respectively. Finally, we use low-level edge information to guide the saliency map generation, and the edge guidance fusion is able to identify saliency regions effectively. The proposed HFFNet has been extensively evaluated on five traditional benchmark datasets. The experimental results demonstrate that the proposed model is fairly effective in salient object detection compared with 10 state-of-the-art models under different evaluation indicators, and it is superior to most of the comparison models. [ABSTRACT FROM AUTHOR]

Data reconstruction, which aims at preserving statistical properties of the data during the reconstruction has become a new criterion for feature selection. Although feature selection could benefit from the perspective of data reconstruction, it is unable to exploit other crucial information, namely, graph structure and pairwise constraints. To address previously mentioned deficiency, we propose a novel feature selection approach in this paper, known as unsupervised feature selection via data reconstruction and side information. More specifically, the proposed method takes advantage of the prior knowledge regarding pairwise constraints (side information), the minimization of data reconstruction error, and the graph embedding simultaneously, such that pivotal features are selected with preserving data manifold structure. To obtain the robust solution, a robust loss function is applied to the feature selection problem, which interpolates between $\ell _{1}$ -norm and $\ell _{2}$ -norm. Eventually, extensive experiments are conducted to demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

People in crowd scenes always exhibit consistent behaviors and form collective motions. The analysis of collective motion has motivated a surge of interest in computer vision. Nevertheless, the effort is hampered by the complex nature of collective motions. Considering the fact that collective motions are formed by individuals, this paper proposes a new framework for both quantifying and detecting collective motion by investigating the spatio-temporal behavior of individuals. The main contributions of this work are threefold: 1) an intention-aware model is built to fully capture the intrinsic dynamics of individuals; 2) a structure-based collectiveness measurement is developed to accurately quantify the collective properties of crowds; 3) a multi-stage clustering strategy is formulated to detect both the local and global behavior consistency in crowd scenes. Experiments on real world data sets show that our method is able to handle crowds with various structures and time-varying dynamics. Especially, the proposed method shows nearly 10% improvement over the competitors in terms of NMI, Purity and RI. Its applicability is illustrated in the context of anomaly detection and semantic scene segmentation. [ABSTRACT FROM AUTHOR]

The existing unsupervised feature extraction methods frequently explore low-redundant features by an uncorrelated constraint. However, the constrained models might incur trivial solutions, due to the singularity of scatter matrix triggered by high-dimensional data. In this paper, we propose a regularized regression model with a generalized uncorrelated constraint for feature selection, which leads to three merits: 1) exploring the low-redundant and discriminative features; 2) avoiding the trivial solutions and 3) simplifying the optimization. Besides that, the local cluster structure is achieved via a novel constrained spectral analysis for the unsupervised learning, where Must-Links and Cannot-Links are transformed into a intrinsic graph and a penalty graph respectively, rather than incorporated into a mixed affinity graph. Accordingly, a discriminative and uncorrelated feature selection with constrained spectral analysis (DUCFS) is proposed with adopting $\sigma $ -norm regularization for interpolating between F-norm and $\ell _{2,1}$ -norm. Due to the flexible gradient and global differentiability, our model converges fast. Extensive experiments on benchmark datasets among several state-of-the-art approaches verify the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Graph clustering plays an important role in data mining. Based on an input data graph, data points are partitioned into clusters. However, most existing methods keep the data graph fixed during the clustering procedure, so they are limited to exploit the implied data manifold and highly dependent on the initial graph construction. Inspired by the recent development on manifold learning, this paper proposes an Adaptive Consistency Propagation (ACP) method for graph clustering. In order to utilize the features captured from different perspectives, we further put forward the Multi-view version of the ACP model (MACP). The main contributions are threefold: (1) the manifold structure of input data is sufficiently exploited by propagating the topological connectivities between data points from near to far; (2) the optimal graph for clustering is learned by taking graph learning as a part of the optimization procedure; and (3) the negotiation among the heterogeneous features is captured by the multi-view clustering model. Extensive experiments on real-world datasets validate the effectiveness of the proposed methods on both single- and multi-view clustering, and show their superior performance over the state-of-the-arts. [ABSTRACT FROM AUTHOR]

Non-negative Matrix Factorization (NMF) has been one of the most popular clustering techniques in machine leaning, and involves various real-world applications. Most existing works perform matrix factorization on high-dimensional data directly. However, the intrinsic data structure is always hidden within the low-dimensional subspace. And, the redundant features within the input space may affect the final result adversely. In this paper, a new unsupervised matrix factorization method, Discrimination-aware Projected Matrix Factorization (DPMF), is proposed for data clustering. The main contributions are threefold: (1) The linear discriminant analysis is jointly incorporated into the unsupervised matrix factorization framework, so the clustering can be accomplished in the discriminant subspace. (2) The manifold regularization is introduced to perceive the geometric information, and the ${\ell _{2,1}}$ ℓ 2 , 1 -norm is utilized to improve the robustness. (3) An efficient optimization algorithm is designed to solve the proposed problem with proved convergence. Experimental results on one toy dataset and eight real-world benchmarks show the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Mine fires are becoming a serious issue as the intensity of mining increases, especially in deep mines. Loose coal gob has a hidden ignition location and a high possibility of spontaneous combustion, which makes fire prevention difficult. Therefore, based on the theory of gas seepage and the characteristics of loose coal, a model of air leakage and spontaneous combustion in gob is established in this paper. Using working face #10414 in the Yangliu coal mine as an example, the relationship between the three spontaneous coal combustion (CSC) zones and the three stress zones is analyzed and verified by combining a FLAC3D simulation with field monitoring. In addition, the influence of advancing speed on the CSC is discussed, and suggestions for fire prevention are presented. The results show that the variation in the calorific value of the CSC with increasing degree of looseness of the residual coal in the gob forms an arch‐shape. There is a one‐to‐one relationship between the distribution of the three stress zones and the three CSC zones. In addition, as the advancing speed increases, the contact time between the loose coal body and the air decreases and the possibility of CSC decreases. This study provides a scientific basis for fire prevention and control in mines. [ABSTRACT FROM AUTHOR]

Congestion detection is an important aspect of vehicular management. However, most of the existing algorithms are insufficient for real applications. Traditional features are not discriminative which results in rather poor performance under complex scenarios. The deep features can better represent high-level information, but the training of deep network for regression is difficult. To promote the congestion detection, a robust hierarchical deep learning is proposed for the task. In this method, a deep network is designed for hierarchical semantic feature extraction. Different from traditional deep regression networks, which usually directly utilize mean squared error as loss function, a robust metric learning is employed to effectively train the network. Based on this, multiple networks are combined together to further improve the generalization ability. Extensive experiments are conducted and the proposed model is confirmed to be effective. [ABSTRACT FROM AUTHOR]

Abstract In this paper, the characteristics and mechanism of the surface potential of coal containing gas for different loading modes (uniaxial compression, cyclic and graded loading) are investigated using an experimental surface potential testing system. Coal can produce a surface potential signal during gas adsorption. This signal increases with the adsorption time and exhibits a memory effect; that is, when the adsorbed gas pressure is increased, the surface potential increases. The coal containing gas can produce a surface potential signal based on different loading modes (uniaxial, cyclic, and graded loading), and the surface potential increases with the load. The higher the gas pressure is, the lower the peak stress is but the higher the maximum surface potential is. The surface potential under cyclic loading exhibits a significant memory effect and is lower than that during the stress-holding stage during graded loading. The surface potential caused by the loading and failure of coal containing gas is the result of the joint actions of gas and stress, which induce a free charge, a frictional force and an electrokinetic effect. This study offers theoretically significant and practical results that reflect the microscale process and electrical precursors of coal fracturing. Highlights • Coal can produce surface potential signal in the process of absorbing gas. • The higher the gas pressure, the higher the surface potential peak. • The surface potential of cyclic loading has a significant memory effect. [ABSTRACT FROM AUTHOR]

How to tackle high dimensionality of data effectively and efficiently is still a challenging issue in machine learning. Identifying anomalous objects from given data has a broad range of real-world applications. Although many classical outlier detection or ranking algorithms have been witnessed during the past years, the high-dimensional problem, as well as the size of neighborhood, in outlier detection have not yet attracted sufficient attention. The former may trigger the distance concentration problem that the distances of observations in high-dimensional space tend to be indiscernible, whereas the latter requires appropriate values for parameters, making models high complex and more sensitive. To partially circumvent these problems, especially the high dimensionality, we introduce a concept called local projection score (LPS) to represent deviation degree of an observation to its neighbors. The LPS is obtained from the neighborhood information by the technique of low-rank approximation. The observation with high LPS is a promising candidate of outlier in high probability. Based on this notion, we propose an efficient and effective outlier detection algorithm, which is also robust to the parameter ${k}$ of ${k}$ nearest neighbors. Extensive evaluation experiments conducted on twelve public real-world data sets with five popular outlier detection algorithms show that the performance of the proposed method is competitive and promising. [ABSTRACT FROM AUTHOR]

Remote sensing image registration has been widely applied in military and civilian fields, such as target recognition, visual navigation and change detection. The dynamic changes in the sensing environment and sensors bring differences to feature point detection in amount and quality, which is still a common and intractable challenge for feature-based registration approaches. With such multiple perturbations, the extracted feature points representing the same physical location in space may have different location accuracy. Most existing matching methods focus on recovering the optimal feature correspondences while they ignore the diversities of different points in position, which easily brings the model into a bad local extrema, especially when existing with the outliers and noises. In this paper, we present a novel accuracy-aware registration model for remote sensing. A soft weighting is designed for each sample to preferentially select more reliable sample points. To better estimate the transformation between input images, an optimal sparse approximation is applied to approach the transformation by multiple iterations, which effectively reduces the computation complexity and also improves the accuracy of approximation. Experimental results show that the proposed method outperforms the state-of-the-art approaches in both matching accuracy and correct matches. [ABSTRACT FROM AUTHOR]

Objective: Intracranial arterial calcification (IAC) is a risk factor of ischemic stroke. However, the relationship between IAC patterns and clinical outcome of ischemic stroke remains controversial. We aimed to investigate the correlation between IAC patterns and the effects of reperfusion therapy among acute stroke patients. Methods: Consecutive acute ischemic stroke patients who underwent reperfusion therapy were included. IAC was categorized as intimal or medial. Based on its involvement, IAC was further classified as diffuse or focal. Neurologic dysfunction was assessed by the National Institute of Health stroke scale (NIHSS). Clinical outcome including favorable neurologic outcome (FNO) and early neurologic deterioration (END) were assessed. Results: Of 130 patients, 117 had IAC. Intimal IAC was identified in 74.6% of patients and medial IAC was present in 64.6% of patients. Diffuse IAC was present in 31.5% of patients. All diffuse IACs were medial pattern. Diffuse IAC was associated with higher baseline NIHSS (p = 0.011) and less FNO (p = 0.047). Compared with patients with focal or single diffuse IAC, patients with multiple diffuse IAC had higher baseline NIHSS (p = 0.002) and less FNO (p = 0.024). Multivariable linear regression (p < 0.001) and logistic regression (p = 0.027) suggested that multiple diffuse IAC was associated with higher baseline NIHSS and less FNO. No significant association was found between END and different IAC patterns. Interpretation: Multiple diffuse medial IAC may predict severer neurologic dysfunction and less favorable neurologic outcome after reperfusion therapy in acute stroke patients. [ABSTRACT FROM AUTHOR]

• We proposed a new support plan and conducted field tests. • Broken surrounding rock, low degradation strength, unreasonable support parameters, and uncoupling of support and surrounding rock were the main causes of roadway deformation. • The active–passive coupling support scheme had an amplifying effect. • The deformation of the roadway surrounding rock was effectively controlled and the scheme could ensure the stability of the roadway. The safe mining of Xin'an Coal Mine was challenging because of several problems, including the soft roadway seams, fracture development, fracturing and deformation of the incompact surrounding rock, and the unreasonable support parameters. Through the investigation of these problems and the analysis of the influence of active and passive support on the stability of the roadway combined with the theory, we proposed a new support plan and conducted field tests. The obtained results were as follows: broken surrounding rock, low degradation strength, unreasonable support parameters, and uncoupling of support and surrounding rock were the main causes of roadway deformation. To a certain extent, the passive support could prevent the damage of the surrounding rock, prevent the attenuation of the stress of the rock around the anchor point, and enhance the effect of active support. The active support could pre-support the gravity of the broken surrounding rock, improve the stress environment of the surrounding rock, and increase the self-balancing ability of the surrounding rock structure. It also could control the damage zone of surrounding rock and the expansion of the stress arch to the depth, restrict the range of rock strata that the passive support system needed to control, and play a protective role. The active–passive coupling support scheme had an amplifying effect. In this regard, the active–passive coupling support scheme of "bolt, W steel belt" + "anchor cable, anchor cable beam" + "I-steel shed" was practiced, and the relevant support parameters were determined. The scheme was implemented in the 2301 mining roadway, and the displacement convergence deformation of the two walls and the roof and floor of the roadway was monitored. It was found that the deformation of the roadway surrounding rock was effectively controlled and the scheme could ensure the stability of the roadway. The research results could provide a basis for the stability analysis of large deformation of broken soft coal of mining roadway and also a basis for the optimization of support parameters. [ABSTRACT FROM AUTHOR]

Segmentation of the prostate in magnetic resonance image has become more in need for its assistance to diagnosis and surgical planning of prostate carcinoma. Due to the natural variability of anatomical structures, statistical shape model has been widely applied in medical image segmentation. Robust and distinctive local features are critical for statistical shape model to achieve accurate segmentation results. The scale invariant feature transformation (SIFT) has been employed to capture the information of the local patch surrounding the boundary. However, when SIFT feature being used for segmentation, the scale and variance are not specified with the location of the point of interest. To deal with it, the discriminant analysis in machine learning is introduced to measure the distinctiveness of the learned SIFT features for each landmark directly and to make the scale and variance adaptive to the locations. As the gray values and gradients vary significantly over the boundary of the prostate, separate appearance descriptors are built for each landmark and then optimized. After that, a two stage coarse-to-fine segmentation approach is carried out by incorporating the local shape variations. Finally, the experiments on prostate segmentation from MR image are conducted to verify the efficiency of the proposed algorithms. [ABSTRACT FROM PUBLISHER]

The deformation control of roadways surrounded by rock in the fully mechanized amplification sections of extra-thick coal seams is problematic. To analyze the failure and failure characteristics of a support frame, as well as the deformation and failure processes of the surrounding rock, through theoretical analysis and industrial tests, the deformation and support conditions of a return airway of a fully mechanized caving face in an extra-thick coal seam in the Yangchangwan Coal Mine, in the Ningdong mining, area were examined. Combined with limit equilibrium theory and roadway section size, the width of the coal pillar of the return air roadway at the 130,205 working face was calculated to be 6 m. The layout scheme and implementation parameters of roof blasting pressure relief, coal pillar grouting modification, and bolt (cable) support were designed. Based on the analysis, a "Coal pillar optimization–roof cutting destressing–routing modification–rock bolting" system for surrounding rock control in synergy with the fully enlarged section mining roadway in the extra-thick coal seam was proposed, and the deformation of the surrounding rock was monitored, along with the stress of the support body and the grouting effect on the site. Field experiments show that after the implementation of the surrounding rock control in synergy with the roadway, the maximum subsidence of the top plate was 55 mm, the maximum bottom heave of the bottom plate was 55 mm, the maximum values of the upper and lower side drums were 30 mm and 70 mm, respectively, and the breaking rate of the bolt (cable) and the deformation of the surrounding rock of the roadway was reduced by more than 90% and 70%, respectively. The effective performance of the coal pillar grouting was observed as well. Field practice of the roadway surrounding rock control in the synergy method indicated that rock deformation was effectively controlled, and the successful application of this technology was able to provide reliable technical and theoretical support for the Ningdong mining area and mines with similar conditions. [ABSTRACT FROM AUTHOR]