Search

Deep learning has made remarkable strides in hyperspectral image (HSI) classification, significantly improving classification performance. However, the challenge of obtaining accurately labeled training samples persists, primarily due to the subjectivity of human annotators and their limited domain knowledge. This often results in erroneous labels, commonly referred to as label noise. Such noisy labels can critically impair the performance of deep learning models, making it essential to address this issue. While previous studies focused on label noise filtering and label correction, these approaches often require estimating noise rates and may inadvertently propagate noisy labels to clean labels, especially in scenarios with high noise levels. In this study, we introduce an ensemble network-based distillation (END) method specifically designed to address the challenges posed by label noise in HSI classification. The core idea is to leverage multiple base neural networks to generate an estimated label distribution from the training data. This estimated distribution is then used alongside the ground-truth labels to train the target network effectively. Moreover, we propose a parameter-adaptive loss function that balances the impact of both the estimated and ground-truth label distributions during the training process. Our approach not only simplifies architectural requirements but also integrates seamlessly into existing deep learning frameworks. Comparative experiments on four hyperspectral datasets demonstrate the effectiveness of our method, highlighting its competitive performance in the presence of label noise. [ABSTRACT FROM AUTHOR]

Deep learning-based label noise learning methods provide promising solutions for hyperspectral image (HSI) classification with noisy labels. Currently, label noise learning methods based on deep learning improve their performance by modifying one aspect, such as designing a robust loss function, revamping the network structure, or adding a noise adaptation layer. However, these methods face difficulties in coping with relatively high noise situations. To address this issue, this paper proposes a unified label noise learning framework with a dual-network structure. The goal is to enhance the model's robustness to label noise by utilizing two networks to guide each other. Specifically, to avoid the degeneration of the dual-network training into self-training, the "disagreement" strategy is incorporated with co-learning. Then, the "agreement" strategy is introduced into the model to ensure that the model iterates in the right direction under high noise conditions. To this end, an agreement and disagreement-based co-learning (ADCL) framework is proposed for HSI classification with noisy labels. In addition, a joint loss function consisting of a supervision loss of two networks and a relative loss between two networks is designed for the dual-network structure. Extensive experiments are conducted on three public HSI datasets to demonstrate the robustness of the proposed method to label noise. Specifically, our method obtains the highest overall accuracy of 98.62%, 90.89%, and 99.02% on the three datasets, respectively, which represents an improvement of 2.58%, 2.27%, and 0.86% compared to the second-best method. In future research, the authors suggest using more networks as backbones to implement the ADCL framework. [ABSTRACT FROM AUTHOR]

Ensembles of extreme learning machine (ELM) have been widely used for hyperspectral image classification. The previous studies have shown that the goal of ensemble learning is to train accurate but diverse component classifiers to improve the generalization performance. To approach this goal, this article proposes a novel framework to construct an ELM ensemble model. The proposed framework relies on multiview-based random rotation ensemble pruning (MVRR-EP) and has several novel features. First, to ensure that the subsets of spectral bands can sufficiently learn the target concept, the spectral bands are divided into multiviews by using correlation analysis. Second, random rotation, a new approach of space transformation, is introduced to transform each view into multiple coordinate spaces, which makes the component classifiers trained on the transformed spaces have great diversity. Third, an accuracy guided ensemble pruning strategy is designed for pruning the component classifiers with low complementarity, and consequently, the remaining component classifiers with high complementarity are combined to a construct ensemble classifier. These techniques guarantee that the component classifiers used to construct an ensemble classifier are accurate but diverse, which ultimately improves the performance of the ensemble classifier. To demonstrate the effectiveness of the proposed MVRR-EP, extensive experiments were carried out on four hyperspectral data sets. Experimental results verify that compared with other methods, the proposed method provides competitive results. [ABSTRACT FROM AUTHOR]

Highlights • Injecting liquid KNa alloy into sandwich-like carbon felt at room temperature. • Synthesizing nano-scale SPAN via a one-step facile solid state reaction. • The new battery presents good cycle stability (capacity retention 95.5%/100 cycles). Abstract Considering high earth abundance and low cost of potassium resource, potassium ion batteries are very promising candidates for energy storage devices due to its similar electrochemical performance to lithium ion batteries and more negative potential than sodium anode. K–Na alloy extending from 41.8 to 90.8 wt% potassium is liquid at ambient temperature. It was used as anode for a rechargeable potassium–sulfur battery in this work because of its high capacity and dendrite-free when charging and discharging. Sulfurized polyacrylonitrile (SPAN) nanocomposites were synthesized and adopted as cathode in the novel battery within good performance. The novel battery presented better cycling stability than K–SPAN battery. We obtained a highly reversible capacity of 513 mAh g−1 and excellent cycling performance of 490 mAh g−1 retention after 100 cycles at a current density of 35 mA g−1, in addition the coulombic efficiency approximately approached to 100% in successive cycles. [ABSTRACT FROM AUTHOR]

Highlights • We proposed a weighted heterogeneous distance metric (WHDM). • We presented WHDM and Dempster–Shafer theory based k NN algorithm. • We proposed a multimodal perturbation method (RRSB) for k NN ensemble. • The effectiveness of our algorithms was shown on multiple UCI data sets and a KDD data set. Abstract In this paper, to address the issue that ensembling k -nearest neighbor (k NN) classifiers with resampling approaches cannot generate component classifiers with a large diversity, we consider ensembling k NN through a multimodal perturbation-based method. Since k NN is sensitive to the input attributes, we propose a weighted heterogeneous distance Metric (WHDM). By using a WHDM and evidence theory, a progressive k NN classifier is developed. Based on a progressive k NN, the random subspace method, attribute reduction, and Bagging, a novel algorithm termed RRSB (reduced random subspace-based Bagging) is proposed for construct ensemble classifier, which can increase the diversity of component classifiers without damaging the accuracy of the component classifiers. In detail, RRSB adopts the perturbation on the learning parameter with a weighted heterogeneous distance metric, the perturbation on the input space with random subspace and attribute reduction, the perturbation on the training data with Bagging, and the perturbation on the output target of k neighbors with evidence theory. In the experimental stage, the value of k , the different perturbations on RRSB and the ensemble size are analyzed. In addition, RRSB is compared with other multimodal perturbation-based ensemble algorithms on multiple UCI data sets and a KDD data set. The results from the experiments demonstrate the effectiveness of RRSB for k NN ensembling. [ABSTRACT FROM AUTHOR]

This paper shows the design and modeling of an end effector with a bidirectional telescopic mechanism to allow a surgical assistant robot to hold and handle surgical instruments. It also presents a force-free control algorithm for the direct teaching of end effectors. The bidirectional telescopic mechanism can actively transmit force both upwards and downwards by staggering the wires on both sides. In order to estimate and control torque via motor current without a force/torque sensor, the gravity model and friction model of the device are derived through repeated experiments. The LuGre model is applied to the friction model, and the static and dynamic parameters are obtained using a curve fitting function and a genetic algorithm. Direct teaching control is designed using a force-free control algorithm that compensates for the estimated torque from the motor current for gravity and friction, and then converts it into a position control input. Direct teaching operation sensitivity is verified through hand-guiding experiments. [ABSTRACT FROM AUTHOR]

Deep learning (DL) has recently been a core ingredient in modern computer vision tasks, triggering a wave of revolutions in various fields. The hyperspectral image (HSI) classification task is no exception. A wide range of DL-based methods have shone brilliantly in HSI classification. However, understanding how to better exploit spectral and spatial information regarding HSI is still an open area of enquiry. In this article, we propose a hybrid convolution and hybrid resolution network with double attention for HSI classification. First, densely connected 3D convolutional layers are employed to extract preliminary spatial–spectral features. Second, these coarse features are fed to the hybrid resolution module, which mines the features at multiple scales to obtain high-level semantic information and low-level local information. Finally, we introduce a novel attention mechanism for further feature adjustment and refinement. Extensive experiments are conducted to evaluate our model in a holistic manner. Compared to several popular methods, our approach yields promising results for four datasets. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) play an important role in hyperspectral image (HSI) classification due to their powerful feature extraction ability. Multiscale information is an important means of enhancing the feature representation ability. However, current HSI classification models based on deep learning only use fixed patches as the network input, which may not well reflect the complexity and richness of HSIs. While the existing methods achieve good classification performance for large-scale scenes, the classification of boundary locations and small-scale scenes is still challenging. In addition, dimensional dislocation often exists in the feature fusion process, and the up/downsampling operation for feature alignment may introduce extra noise or result in feature loss. Aiming at the above issues, this paper deeply explores multiscale features, proposes an adaptive attention constraint fusion module for different scale features, and designs a semantic feature enhancement module for high-dimensional features. First, HSI data of two different spatial scales are fed into the model. For the two inputs, we upsample them using bilinear interpolation to obtain their subpixel data. The proposed multiscale feature extraction module is intended to extract the features of the above four parts of the data. For the extracted features, the multiscale attention fusion module is used for feature fusion, and then, the fused features are fed into the high-level feature semantic enhancement module. Finally, based on the fully connected layer and softmax layer, the prediction results of the proposed model are obtained. Experimental results on four public HSI databases verify that the proposed method outperforms several state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs), a kind of feedforward neural network with a deep structure, are one of the representative methods in hyperspectral image (HSI) classification. However, redundant information and interclass interference are common and challenging problems in HSI classification. In addition, if the spectral and spatial information is not properly extracted and analyzed, it will affect the classification performance of the network to a great extent. Aiming at these issues, this article proposes an HSI classification method based on an adaptive hash attention mechanism and a lower triangular network (AHA-LT). First, the attention mechanism is introduced in the preprocessing stage, which is composed of the spectral attention module and the adaptive hash spatial attention module in series. Then, the data processed by the attention mechanism are introduced into the lower triangular network (LTNet) to obtain the fused high-dimensional semantic features. Finally, we compress the features and obtain the output classification results through several fully connected layers. Among them, LTNet is composed of 2-D–3-D CNN and multiscale features. The network integrates the characteristics of multibranch, feature fusion, feature compression, and skip connections. Extensive experiments on four widely used HSI data sets show that the proposed method can obtain a great improvement in performance compared with the existing methods. [ABSTRACT FROM AUTHOR]

Random forest (RF) has obtained great success in hyperspectral image (HSI) classification. However, RF cannot leverage its full potential in the case of limited labeled samples. To address this issue, we propose a unified framework that embeds active learning (AL) and semi-supervised learning (SSL) into RF (ASSRF). Our aim is to utilize AL and SSL simultaneously to improve the performance of RF. The objective of the proposed method is to use a small number of manually labeled samples to train classifiers with relative high classification accuracy. To achieve this goal, a new query function is designed to query the most informative samples for manual labeling, and a new pseudolabeling strategy is introduced to select some samples for pseudolabeling. Compared with other AL- and SSL-based methods, the proposed method has several advantages. First, ASSRF utilizes the spatial information to construct a query function for AL, which can select more informative samples. Second, in addition to providing more labeled samples for SSL, the proposed pseudolabeling method avoids bias caused by AL-labeled samples. Finally, the proposed model retains the advantages of RF. To demonstrate the effectiveness of ASSRF, we conducted experiments on three real hyperspectral data sets. The experimental results have shown that our proposed method outperforms other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Sulfurized polyacrylonitrile (SPAN) nanocomposites were synthesized and used as a cathode in a novel rechargeable Na+/K+ hybrid battery with high performance for the first time. When 0.9 mol NaPF6 and 0.1 mol KPF6 were dissolved in ethylene carbonate (EC)/dimethyl carbonate(DMC)/ethyl methyl cabonate(EMC) (4:3:2, v/v/v), used as hybrid electrolyte, Na foil was used as the anode, and SPAN composites were used as the cathode, a hybrid ion system was created via composition–decomposition between Na+/K+ and SPAN and stripping–depositing of Na+ with suppressed dendrites by taking advantage of the self-healing electrostatic shield effect. As a result, a highly reversible calculated capacity of 1405.5 mAh gsulfur−1 with a coulombic efficiency approaching 100% after 100 cycles was obtained at a current density of 35 mA g−1. This environmentally benign, low-cost Na+/K+ hybrid battery shows promise as a new future flexible energy storage system (ESS) technology. [ABSTRACT FROM AUTHOR]

Recently, convolutional neural network (CNN) and Transformer have been successfully applied in hyperspectral image (HSI) classification. Despite their success, some challenges persist. Popular HSI classification methods typically employ only a single patch as the network input, which may not provide robust information for various datasets or within different scenarios of a specific dataset. Large-scale patches are usually conducive to classifying large-scale scenes, while small-scale patches excel at classifying boundary locations. In addition, the majority of methods do not distinguish between the central pixel and its adjacent pixels, although the information from the central position is crucial. In this paper, a novel double-branch hierarchical partial convolution network (DBHP-Net) is proposed to address the aforementioned challenges. Concretely, DBHP-Net employs a new convolution, named TConv, to emphasize mining the spatial-channel features of the central position. Then, the TConvFormer Block is built by adopting the advanced engineering techniques of Transformer. Furthermore, we construct our model in a hierarchical style. The double-branch structure with large-scale and small-scale patches provides complementary information from the dataset. Experimental results based on randomly selected samples and spatially disjoint samples of two benchmark datasets demonstrate that our method outperforms other state-of-the-art (SOTA) methods. [ABSTRACT FROM AUTHOR]

As a foundation task of natural language processing, text classification is widely used in information retrieval, public opinion analysis, and other related tasks. Facing the problem of sparse features of Chinese short texts, which affects the classification accuracy of Chinese short texts, this paper proposes a Chinese short text classification method based on the Character Frequency Sub-word Enhancement (CFSE), which can effectively improve the classification accuracy of Chinese short texts. First, the initial Chinese-character sequence is mapped to the corresponding Character Frequency Sub-word (CFS) sequence based on the global character1 frequency information. Second, the relationship features among data are extracted based on BiLSTM-Att processing CFS sequence, and the semantic features of the initial Chinese-character sequence are obtained through ERNIE. Finally, these two kinds of features are fused and input into the text classifier to obtain the classification results. Experimental results show that the proposed method can improve the classification accuracy of Chinese short texts. [ABSTRACT FROM AUTHOR]

Sparse representation theory shows effectiveness in single image super-resolution (SR). Existing image super-resolution methods usually make use of l 1 -regularization, l 2 -regularization or their combination to restrict the sparsity. However, the nonlocal similarity of images, which can be helpful to image SR, is often neglected. In order to utilize the nonlocal similarity and improve SR results in this paper, we propose a new single image super-resolution method by combining the adaptive sparse representation and robust principal component analysis (RPCA). Furthermore, we adopt the self-similarity learning framework to construct the dictionary pair. In our method, we first compute the sparse coefficient of each testing image patch through adaptive sparse representation with the constructed dictionary. Then, for each testing image block, we search for its similar patches and use RPCA as a low-rank optimization strategy to the corresponding coefficients. Extensive experiment results demonstrate that the proposed method can possesses better performance compared with some state-of-the-art methods. [ABSTRACT FROM AUTHOR]

At present, most high-performance cellulose matrix composites only use cellulose as reinforcement material, which is an obstacle to maximize the advantages of nanocellulose in structure and properties. The development of new functional nanocomposites with cellulose as the main component can better meet people's needs for high-performance and degradable composites, which requires a comprehensive and thorough understanding of cellulose. Considering the limitations of physical experiments, we performed molecular dynamics simulation of the uniaxial tensile behavior of the cellulose system at three different strain rates (10−4/ps, 10−5/ps, and 10−6/ps), and the stress-strain responses of cellulose systems at different strain rates are obtained. The effect of the strain rate on the mechanical properties of amorphous cellulose system during the tensile processes is analyzed. The deformation mechanism of cellulose amorphous system during the tensile processes is characterized by the energy changes of the different terms including dihedral angle torsion term, bond tensile term, angle bending term, and nonbond term. Structural evolution of the cellulose crystal system during the tensile processes is used to explain the failure mechanism of cellulose. The kinetic simulation results show that the mechanical properties of the cellulose amorphous system increase with the increase of strain rate. Compared with the strain rate of 10−5/ps, the elastic modulus of the system increases by 6.73 GPa at the strain rate 10−4/ps. During the tensile processes, cellulose amorphous region adapts to the applied load mainly through the stretching of the cellulose macromolecular chains, i.e., the deformation of bond lengths and bond angles, without any breakage of the molecular chains. The main causes of chain lengthening at different strain rates are different. The failure of cellulose is caused by the slip and rearrangement of some molecular chains in the crystal structure. [ABSTRACT FROM AUTHOR]

The hinged connections are widely used in agricultural machinery, such as the three-point hitch device at the rear of the tractor and the header position adjustment device of the corn harvester. The hinged connections in agricultural machinery allow relative rotations between the connected parts so that the positions of the agricultural equipment can be adjusted according to the operation requirements. Thus, the hinged connections are always subject to large and complex forces during operation, leading to complicated nonlinear vibrations, which significantly impact the operational performance and fatigue life of the agricultural machinery. Therefore, the nonlinear dynamics characteristics of the hinged connections in agricultural machinery are studied in this paper. Firstly, the parts connected by the hinge are simplified to the hinged beam structures according to the mechanical structure characteristics and loading features of the hinged connection structures in agricultural machinery. Then, the equivalent dynamics model of the hinged beam structure is established based on the nonlinear contact force model at the hinged connection and the finite element method. Finally, the effects of parameters such as hinge clearances, excitation frequencies, excitation amplitudes and hinge contact stiffness on the nonlinear dynamics characteristics of the hinged beam structures are analyzed by solving the dynamics responses of the system using numerical simulations. The research results can provide a reference for the optimal design of the hinged connection structures. [ABSTRACT FROM AUTHOR]

As a foundation task of natural language processing, text classification is widely used in information retrieval, public opinion analysis, and other related tasks. Facing the problem of sparse features of Chinese short texts, which affects the classification accuracy of Chinese short texts, this paper proposes a Chinese short text classification method based on the Character Frequency Sub-word Enhancement (CFSE), which can effectively improve the classification accuracy of Chinese short texts. First, the initial Chinese-character sequence is mapped to the corresponding Character Frequency Sub-word (CFS) sequence based on the global character1 frequency information. Second, the relationship features among data are extracted based on BiLSTM-Att processing CFS sequence, and the semantic features of the initial Chinese-character sequence are obtained through ERNIE. Finally, these two kinds of features are fused and input into the text classifier to obtain the classification results. Experimental results show that the proposed method can improve the classification accuracy of Chinese short texts. [ABSTRACT FROM AUTHOR]

The difficulty of feature extraction and the small sample size are two challenges in the field of mechanical fault diagnosis for a long time. Here we propose an intelligent mechanical fault diagnosis method for scenario with small sample datasets. This method can not only diagnose bearing faults but also gear faults, and has strong generalization performance. We use convolutional neural network to realize automatic feature extraction. Through sliding window scanning, one sample set is expanded to three sub-sample sets with different scales to meet the needs of deep learning training. Three convolutional networks are used to extract the features of the subsets respectively to ensure that their useful features are fully extracted. After feature extraction, the feature is reconstructed through feature splicing. Because of the unique advantages of SVM in dealing with small sample sets, we use SVM to classify the reconstructed features. We use the bearing data set collected by Case Western Reserve University in the United States, the bearing fault data set collected by Xi'an Jiaotong University in China, and the gearbox fault data collected by the University of Connecticut in the United States to conduct experiments. The experimental results show that the accuracy of training, validation and testing of the proposed method on the three data sets all reach 100%. This proves that our method can not only tackle the two challenges, but also has high fault diagnosis accuracy and strong generalization performance. It is hoped that our proposed method can contribute to the development of mechanical fault diagnosis. [ABSTRACT FROM AUTHOR]

For the problem of friction damage in cotton fiber processing, a multi-scale combination of investigation methods is proposed. The surface of damaged cotton fiber is detected by relevant test means with damage features such as dislocations, defects and cracks. The internal pyranose ring and glycosidic bond fail, the crystallinity decreases, and the number of hydrogen bonds decreases. Anisotropy exists in the frictional properties of the microscopic surface of the cotton fiber. The results of molecular dynamics simulation showed that the cellulose main chain failed mainly at the glycosidic bond, and the side chain failed mainly at the hydroxymethyl functional group. Its interchain hydrogen bond O 3 H...O 5 was the least damaged. The cellulose crystal (200) surfaces had poor abrasion resistance, and the frictional properties of each crystal surface were anisotropic. The results of the study provide a theoretical basis for improving friction and wear problems in cotton fiber processing. [ABSTRACT FROM AUTHOR]

For the problem of friction damage in cotton fiber processing, a multi-scale combination of investigation methods is proposed. The surface of damaged cotton fiber is detected by relevant test means with damage features such as dislocations, defects and cracks. The internal pyranose ring and glycosidic bond fail, the crystallinity decreases, and the number of hydrogen bonds decreases. Anisotropy exists in the frictional properties of the microscopic surface of the cotton fiber. The results of molecular dynamics simulation showed that the cellulose main chain failed mainly at the glycosidic bond, and the side chain failed mainly at the hydroxymethyl functional group. Its interchain hydrogen bond O 3 H...O 5 was the least damaged. The cellulose crystal (200) surfaces had poor abrasion resistance, and the frictional properties of each crystal surface were anisotropic. The results of the study provide a theoretical basis for improving friction and wear problems in cotton fiber processing. [ABSTRACT FROM AUTHOR]

This study investigated the frictional contact behavior of the plain cotton fabric with the metal through linear friction experiments and real contact analysis. It established a theoretical calculation model of the contact area based on the characteristics of contact behavior. The results indicate that the contact between the cotton fabric and the metal is a multi-point contact, with the contact occurring on the fabric knuckles. The contact behavior is mainly influenced by the normal load, fabric knuckle size, and the number of fabric knuckles. The theoretical calculation model of the contact area considers the multi-scale structural characteristics of cotton fabrics and validates the applicability of the friction adhesion theory to cotton fabrics. The effects of normal load and velocity on the tribological characteristics of cotton fabrics are explained based on the number of contact points. [ABSTRACT FROM AUTHOR]

This study developed three friction models including the friction between crystalline cellulose and crystalline cellulose (FMCC), crystalline cellulose and amorphous cellulose (FMCA), and amorphous cellulose and amorphous cellulose (FMAA) to analyze the effects of loading and sliding velocity on friction and wear using molecular dynamics simulation. It is found that the friction of FMCC consists of three stages, that is the stage of static to kinetic friction (S I), the kinetic friction to stable friction (S II), and the stable friction (S III), however, the FMCA and FMAA without the S II. The velocity affects the time to enter the S III significantly, and the loading influences the friction in the S III significantly. And amorphous cellulose is more susceptible to wear than crystalline cellulose. [ABSTRACT FROM AUTHOR]

• The gradual transition of the plowshare from the initial wear stage to the stable wear stage. • Changes in wear size are concentrated at the edge of the plowshare. • Dp value is less than 0.2, the wear mechanism is micro-plowing. • The coupling of abrasive wear, impact wear and corrosive wear exacerbates the wear of the plowshare. The ploughshare represents a pivotal component of high-speed plough. The condition of the ploughshare's wear and tear directly affects a soil tilling operation. This paper carries out experimental field wear studies on the ploughshare of a high-speed plough to investigate the changes in mass, thickness, cutting edge parameters, wear profile, 3D wear morphology, microscopic wear morphology, and tissue composition of the 65Mn ploughshare in 35 ha, 70 ha, 105 ha, and 140 ha cultivation areas, and calculates the degree of penetration (D p value) of the wear surface to analyze the wear mechanism, with a view to prolonging the service life of the ploughshare from the perspective of its design and materials. The results showed that the ploughshare showed adaptive wear changes in response to cultivation, most pronounced in the first 35 ha. The wear rate of the ploughshare shows a decreasing trend with the increase of the cultivated area, indicating a gradual transition from the initial wear stage to the stable wear stage. During cultivation, the stresses are concentrated in the ploughshare cutting edge area, and the wear size of the cutting edge varies significantly. Wear of the ploughshare in parts other than the cutting edge is dominated by a reduction in thickness. The roughness of the wear surface stabilized at about 0.4 μm throughout the wear period. With the change of Y-axis position, the wear mechanism changes from impact wear to micro-ploughing, and finally the form of micro-ploughing action decreases gradually. The D p values of the calculated wear surface grooves are all less than 0.2, and the wear mechanism is micro-plowing. Observation of the microscopic wear morphology reveals traces of corrosive wear in addition to abrasive wear, where the soil reacts chemically with the metal and cracks the surface and subsurface layers of the metal. The coupling of abrasive wear, impact wear and corrosive wear exacerbates the wear of the ploughshare. [ABSTRACT FROM AUTHOR]

• Performance characterization of electroplated chromium coating during dry friction process. • Surface crack propagation evolution process of electroplated chromium coating. • Mechanical analysis of surface crack propagation. Electroplated chromium coatings are commonly used to improve the wear resistance of metal parts in the cotton textile industry, and surface crack is an important feature. This study investigates the tribological properties and surface crack propagation of the electroplated chromium coating in dry friction contact with cotton fabric. It is found that the coating has excellent tribological properties, with the friction coefficient eventually stabilized at 0.11 and the average wear rate of 0.0038 mm3/h. The cotton fabric mainly exhibits a polishing effect on the coating during friction, and the failure of the coating is in the form of delamination failure within the coating. Crack propagation is mainly due to the friction of cotton fabric, the main cracks of the mesh crack structure hardly propagate, while the short cracks extending on the main cracks undergo propagation behavior. Short cracks propagate vertically along the friction direction and eventually develop into the main crack level. The tangential stress generated by friction acts on the crack tip, which is the main cause of crack propagation. The plastic zone generated at the crack tip affects the distribution of the stress field and the effect of the plastic zone on crack propagation is analyzed. [ABSTRACT FROM AUTHOR]

The effects of load and temperature on the friction between crystalline cotton cellulose and chromium in vacuum were investigated utilizing ReaxFF molecular dynamics. Simulation results indicate that a new chemical bond, Cr=O bond, is formed between the sliding friction interface. In the initial stage, the friction force is determined by the number of atoms in contact with the interface. Then, the friction force depends on the number of atoms of cellulose nested in the chromium matrix at the contact interface. It is positively correlated with load and temperature. Under low load, with the formation of the Cr=O bond, the surface structure of the chromium matrix is damaged. With the increase of load, more Cr=O bonds are formed between the contact interfaces, which leads to more profound damage to the surface structure of the chromium layer. This work systematically introduces the influence mechanism of load and temperature on the friction behavior of crystalline cotton cellulose with chromium, thus providing a new perspective on the study of frictional wear mechanism between cotton cellulose and metal. [ABSTRACT FROM AUTHOR]

Sparse coding based-methods show great effectiveness in single image super-resolution (SR). Existing methods generally use only synthesis sparse coding. However, the analysis sparse coding model, which is an alternative to the synthesis sparse coding model, is typically neglected In this paper, we propose a novel single image SR method by combining synthesis sparse coding with analysis sparse coding. In contrast to the existing sparse coding-based SR methods, we replace synthesis sparse coding with analysis sparse coding in the low-resolution (LR) feature representation phase. Thus, we use an analysis dictionary to obtain the LR coefficients and use a synthesis dictionary to reconstruct the high-resolution (HR) patches. Due to the introducing of analysis sparse representation, the l 0 or l 1 optimization is replaced by soft threshold shrinkage, which is a more time-saving method. To improve the convergence of the training model, we introduce a linear mapping function that reveals the relationship between the HR coefficients and LR coefficients. An alternating optimization strategy is adopted to solve the improved training model. Furthermore, global and nonlocal constraints are taken into account to improve the quality of the reconstructed images. Compared with some existing SR methods, the experimental results demonstrate that our proposed method can obtain better performance. • We first introduce synthesis and analysis sparse coding simultaneously into a training model for single image SR problem. • The introduction of analysis sparse coding model can reduce the time consumption in reconstruction stage. • We introduce a linear mapping function to reveal the relationship between them. • The proposed method can possess competitive performance compared with some state-of-art methods. [ABSTRACT FROM AUTHOR]

This paper presents a novel method for accurate people counting in highly dense crowd images. The proposed method consists of three modules: extracting foreground regions (EF), pixel-wise attention mechanism (PAM) and single-column density map estimator (S-DME). EF can suppress the disturbance of complex background efficiently with a fully convolutional network, PAM performs pixel-wise classification of crowd images to generate high-quality local crowd density maps, and S-DME is a carefully designed single-column network that can learn more representative features with much fewer parameters. In addition, two new evaluation metrics are introduced to get a comprehensive understanding of the performance of different modules in our algorithm. Experiments demonstrate that our approach can get the state-of-the-art results on several challenging datasets including our dataset with highly cluttered environments and various camera perspectives. [ABSTRACT FROM AUTHOR]

This study employs a combination of experiments and molecular dynamics to analyze the mechanical properties and surface damage characteristics of cotton fibers during the combing process. Additionally, it investigates the alterations in physical and chemical properties at the atomic scale resulting from mechanical damage. Raw cotton (RC) is combed to 1st combed cotton (1st CC), 2nd combed cotton (2nd CC) and 3rd combed cotton (3rd CC). It was found that the mechanical properties and crystallinity showed an increasing and then decreasing trend with the process of combing, and the degree of surface tearing increased, and the binding energy of C and O shifted to a lower position. The breaking strength of cotton fibers first increased by 7.4 % and then decreased by 11 % and 7.7 % respectively, and the crystallinity was CrI (RC) = 70.8 %, CrI (1st CC) = 75.3 %, CrI (2nd CC) = 72.7 %, and CrI (3rd CC) = 71.8 % respectively. The C–O bond and the C–C bond at the amorphous regions are broken after combing lead to the cellulose chain to break, resulting in a decrease in the breaking strength of the fibers. The C–O bond as well as the C-O-C bond angles changes significantly during stretching, and the increase in ordering of the amorphous regions causes an increase in crystallinity. [ABSTRACT FROM AUTHOR]

Rechargeable lithium-sulfur (Li-S) batteries are considered as one of the most potential energy storage devices owing to their high theoretical capacity, environmental friendliness and low cost. However, the practical applications of Li-S batteries are restricted by the low conductivity of sulfur, serious dissolution of intermediate polysulfides and fast capacity degradation. Herein, we design a titanium nitride (TiN) modified separator to improve the long-term cycle stability and discharge capacity of Li-S batteries. Strong interaction between the TiN particles and polysulfides can greatly restrain the “shuttle effect” and benefit the cycle stability. With this modified separator, cathodes with a sulfur loading of 78.1% could deliver a high initial discharge capacity of 1296 mAh/g and demonstrate a decay rate of 0.17% per cycle under a high current density of 0.5C. Such superior electrochemical performances ascribe to the high conductivity of TiN and strong affinity between TiN particles and polysulfides. [ABSTRACT FROM AUTHOR]

• To suppress the generation of artifact and residual noise in the results of EPLL model, we propose a novel BASMM-EPLL method with superior performance in structure preserving and artifact reducing. • Based on the anisotropic nonlocal self-similarity, we design a content-aware regularization parameter selection scheme to enhance the robustness of the proposed model. • Our model is tested on 16 widely used images and 1 dataset, and achieves an outstanding performance in noise reduction and detail preservation. Meanwhile, the undesirable artifacts and residual noise are suppressed effectively. The expected patch log-likelihood (EPLL) model is a patch prior-based image restoration method which received extensive attention in image processing in recent years for its outstanding ability to preserve the detail and structure. However, due to using the Gaussian mixture model (GMM) with the noise sensitivity as the local prior, the EPLL model suffers from undesired artifact and poor robustness frequently. In this paper, to restrain the generation of artifact of EPLL model, we replace the GMM with a bounded asymmetrical Student's-t mixture model (BASMM), which is sufficiently flexible to fit different shapes of image data, such as non-Gaussian, non-symmetric, and bounded support data. Then, the anisotropic nonlocal self-similarity (ANSS) based regularization parameters are designed to improve the robustness of the proposed model. Experimental results demonstrate the competitiveness of our proposed model compared with that of state-of-the-art methods in performance both visually and quantitatively. [ABSTRACT FROM AUTHOR]

Recently, hyperspectral image (HSI) classification has become a popular research direction in remote sensing. The emergence of convolutional neural networks (CNNs) has greatly promoted the development of this field and demonstrated excellent classification performance. However, due to the particularity of HSIs, redundant information and limited samples pose huge challenges for extracting strong discriminative features. In addition, addressing how to fully mine the internal correlation of the data or features based on the existing model is also crucial in improving classification performance. To overcome the above limitations, this work presents a strong feature extraction neural network with an attention mechanism. Firstly, the original HSI is weighted by means of the hybrid spectral–spatial attention mechanism. Then, the data are input into a spectral feature extraction branch and a spatial feature extraction branch, composed of multiscale feature extraction modules and weak dense feature extraction modules, to extract high-level semantic features. These two features are compressed and fused using the global average pooling and concat approaches. Finally, the classification results are obtained by using two fully connected layers and one Softmax layer. A performance comparison shows the enhanced classification performance of the proposed model compared to the current state of the art on three public datasets. [ABSTRACT FROM AUTHOR]

Deep-learning methods, especially convolutional neural networks (CNN), have become the first choice for hyperspectral image (HSI) classification to date. It is a common procedure that small cubes are cropped from hyperspectral images and then fed into CNNs. However, standard CNNs find it difficult to extract discriminative spectral–spatial features. How to obtain finer spectral–spatial features to improve the classification performance is now a hot topic of research. In this regard, the attention mechanism, which has achieved excellent performance in other computer vision, holds the exciting prospect. In this paper, we propose a double-branch network consisting of a novel convolution named pyramidal convolution (PyConv) and an iterative attention mechanism. Each branch concentrates on exploiting spectral or spatial features with different PyConvs, supplemented by the attention module for refining the feature map. Experimental results demonstrate that our model can yield competitive performance compared to other state-of-the-art models. [ABSTRACT FROM AUTHOR]