Search

A chirp compensation method of fiber Bragg grating (FBG) is proposed to improve the residual stress measurement accuracy in the hole-drilling method. The effects of the radii of the hole, the distance between the FBG and the hole, the Young’s modulus of the material, and the residual stress on the FBG chirped spectra are analyzed. The relationship between radii of the hole, the distance between the FBG and the hole, the full-width at half maximum (FWHM), and the peak wavelength are established. The feasibility of the method is verified through experiments, which shows that the residual stress measurement error can be reduced by the compensation method.

Segmentation of the liver from abdominal CT images is an essential step for computer-aided diagnosis and surgery planning. The U-Net architecture is one of the most well-known CNN architectures which achieved remarkable successes in both medical and biological image segmentation domain. However, it does not perform well when the target area is small or partitioned. In this paper, we propose a novel architecture, called dense feature selection U-Net (DFS U-Net), which addresses this challenging problem. Specifically, The Hounsfield unit values were windowed in a range to exclude irrelevant organs, and then use the pre-processed data to train our proposed DFS U-Net model. To further improve the segmentation accuracy of the small region and disconnected regions of interests with limited training datasets, we improve the loss function by adding a parameter to the formula. With respect to the ground truth, the Dice score ratio can reach over 94.9% for the liver. Our experimental results demonstrate its potential in clinical usage with high effectiveness, robustness and efficiency.

With the expansion of the scale of coal mining, the safety problems caused by the reburning of coal are becoming more and more serious. In this paper, the pyrolysis characteristics of gas coal and the exothermic characteristics of reoxidation of residues were studied by using a synchronous thermal analyzer. The functional groups of pyrolysis residues were tested, and the group content and characteristic structural parameters were calculated based on quantum chemistry method. The results show that with the increase of pyrolysis temperature, Volatile maximum separation rate (Vmax) and the change in the residual weight of the coal sample (ΔWvp) increase. The increase of temperature will lead to the decrease of hydroxyl and aliphatic hydrocarbon content in coal, and the increase of aromatic hydrocarbon. With the deepening of pyrolysis, the ignition point temperature of coal samples decreases first and then rises, the combustion intensity and combustion concentration are strengthened. The pyrolysis results show that 462.8°C is the critical temperature for the transition during pyrolysis. The ignition point of the residue is less affected by the pyrolysis conditions, and the ignition temperature of the raw coal and the pyrolysis residue varies within 330.57°C–334.98°C.

Despite some interesting results from different research groups, a public database for Uyghur online handwriting recognition and a baseline study are not yet available for comparison purpose. In order to fill this void, we present a database of Uyghur online handwritten words and carry out the first benchmark experiments using it. This database contains 125,020 samples of 2030 words collected from 393 writers. According to Uyghur lexicon characteristics, two out-of-vocabulary datasets are especially provided for evaluation. We carry out some unconstrained handwritten word recognition experiments on the database using recurrent neural networks as base model. Recognition results are acquired using connectionist temporal classification without lexicon search and external language model. Concatenated and averaged bidirectional recurrent layers are compared for better generalization. Based on Uyghur unicode representation, we are interested in comparing the models using different alphabets, based both on character types and character forms. To improve generalization, we propose 1D convolutional model which implements 1D convolutional layers for sequence feature extraction. In our experiments, the proposed 1D convolutional model and its variations surpassed the base recurrent layered model on the out-of-vocabulary words by clear margin. 83.23% CAR (character accurate rate) was resulted when out-of-vocabulary samples are used for testing. The highest recognition rate is as high as 94.95% CAR when the test set shares the same lexicon to the training set. The experiments in this paper can be the baseline references for the future study using this database.

Social network information search, especially for microblog search, has been one of the research hotspots in the domain of information search. For complexities of microblog data on arbitrary typing and semantic ambiguity, classical approaches cannot be directly adopted. In this paper, we propose a security topics related microblogs search model based on deep convolutional neural networks (DCNN-CSTRS) to search microblogs similar to a specific security topic contents. This method is trained to capture local semantic features of short microblog texts to filter security topics related contents from microblogs. A matching model based on deep convolution neural network is designed to rank the results by matching the extracted local features of queries and documents respectively through non-linear feature transformations of the convolution and pooling. The matching model ranks the pairs of query-document by similarities. Experimental results demonstrate that the proposed approach performs better compared with the state-of-the-art methods.

As data volumes grow rapidly, distributed computations are widely employed in data-centers to provide cheap and efficient methods to process large-scale parallel datasets. Various computation models have been proposed to improve the abstraction of distributed datasets and hide the details of parallelism. However, most of them follow the single-layer partitioning method, which limits developers to express a multi-level partitioning operation succinctly. To overcome the problem, we present the NDD (Nested Distributed Dataset) data model. It is a more compact and expressive extension of Spark RDD (Resilient Distributed Dataset), in order to remove the burden on developers to manually write the logic for multi-level partitioning cases. Base on the NDD model, we develop an open-source framework called Bigflow, which serves as an optimization layer over computation engines from most widely used processing frameworks. With the help of Bigflow, some advanced optimization techniques, which may only be applied by experienced programmers manually, are enabled automatically in a distributed data processing job. Currently, Bigflow is processing about 3 PB data volumes daily in the data-centers of Baidu. According to customer experience, it can significantly save code length and improve performance over the intuitive programming style.

The C4.5 algorithm has three shortcomings: the wide range of candidate segmentation threshold sequences for continuous attributes, the comprehensive influence of different attributes and local subsets under the same attribute, and the inter-attribute redundancy. When dealing with continuous attributes, sampling and threshold supplement processing near the transition boundary of the attribute interval corresponding to the adjacent different categories are performed for narrowing the range of candate segmentation threshold sequences. By adding standardizing Euclidean distance of the attribute global and local factors to represent attribute weight, the calculation of C4.5 information gain is otpimized. And averaging Gini index of other attributes and adding correction factor, the influence of redundancy between attributes is greatly decreased. The overall average improvement range of the base classifier and the bagging integration classifier is 0.6%~2.1% and 0.7% ~ 2.7%, respectively, which shows that this integration model can improve the classification accuracy and also validate its feasibility and reliability.

Accurate pancreas segmentation from medical images is an important yet challenging problem for medical image analysis and medical surgery. Challenges relating to the pancreas image acquisition, the limited availability of image data and segmentation methodology hinder this segmentation task. This paper aims to present a systematic review of the different methodologies. Recently, a variety of segmentation methods have been proposed for automatic delineation of pancreas images. We intended to survey the methods proposed for image segmentation, outline those that have already been adopted for pancreas segmentation, perform a comparison between them using various indices and experimental data, and discuss their major contributions. Looking at the theoretical approach, different segmentation methods have been applied for delineating the pancreas and can be classified into those based on region, edge, atlas, neural network and other categories. Each segmentation method has its own advantages and disadvantages. Based on the previously performed analysis and discussion, and compared to other abdominal organs, improving the results of pancreas segmentation in the near future will require addressing several challenging issues. These issues include: (1) establishing publically available standardized datasets; (2) define clear metrics to evaluate the segmentation performance and (3) design new segmentation methods.

Defects can induce drastic changes of the electronic properties of two-dimensional transition metal dichalcogenides and influence their applications. It is still a great challenge to characterize small defects and correlate their structures with properties. Here, we show that tip-enhanced Raman spectroscopy (TERS) can obtain distinctly different Raman features of edge defects in atomically thin MoS2, which allows us to probe their unique electronic properties and identify defect types (e.g., armchair and zigzag edges) in ambient. We observed an edge-induced Raman peak (396 cm−1) activated by the double resonance Raman scattering (DRRS) process and revealed electron–phonon interaction in edges. We further visualize the edge-induced band bending region by using this DRRS peak and electronic transition region using the electron density-sensitive Raman peak at 406 cm−1. The power of TERS demonstrated in MoS2 can also be extended to other 2D materials, which may guide the defect engineering for desired properties. Probing inevitable defects in two- dimensional materials is challenging. Here, the authors tackle this issue by using tip-enhanced Raman spectroscopy (TERS) to obtain distinctly different Raman features of edge defects in atomically thin MoS2, and further probe their unique electronic properties as well as identify the armchair and zigzag edges.

Objective To evaluate the clinical effect differences between conducting qi and regulating the spirit needling method and regular needling method. Methods Sixty-nine patients of insomnia were randomly divided into an observation group (35 cases) and a control group (34 cases). Baihui (百会 GV20), Sishencōng (四神聪 EX-HN1), Hegŭ (合谷 LI4), Taichōng (太冲 LR3), Zusānlĭ (足三里 ST36) and Sānyīnjiāo (三阴交 SP6) were selected in the two groups. The patients in the observation group were treated with needling method for conducting qi and regulating the spirit. The patients in the control group were treated with the regular needling method. The treatment was given once every other day, both groups were treated for 12 times. Pittsburgh Sleep Quality Index (PSQI) and Epworth Sleepiness Scale (ESS) were observed before and after the treatment and the clinical effect was evaluated. Results After treatment, the total score and the score of each item of PSQI as well as the result of ESS were all improved after the treatment in the two groups (all P Conclusion The conducting qi and regulating the spirit needling method achieves a better effect on insomnia as compared with the regular needling method. This method achieves a remarkable improvement in sleep latency, sleep quality and daytime sleepiness.

Transductive zero-shot learning (TZSL) extends conventional ZSL by leveraging (unlabeled) unseen images for model training. A typical method for ZSL involves learning embedding weights from the feature space to the semantic space. However, the learned weights in most existing methods are dominated by seen images, and can thus not be adapted to unseen images very well. In this paper, to align the (embedding) weights for better knowledge transfer between seen/unseen classes, we propose the virtual mainstay alignment network (VMAN), which is tailored for the transductive ZSL task. Specifically, VMAN is casted as a tied encoder-decoder net, thus only one linear mapping weights need to be learned. To explicitly learn the weights in VMAN, for the first time in ZSL, we propose to generate virtual mainstay (VM) samples for each seen class, which serve as new training data and can prevent the weights from being shifted to seen images, to some extent. Moreover, a weighted reconstruction scheme is proposed and incorporated into the model training phase, in both the semantic/feature spaces. In this way, the manifold relationships of the VM samples are well preserved. To further align the weights to adapt to more unseen images, a novel instance-category matching regularization is proposed for model re-training. VMAN is thus modeled as a nested minimization problem and is solved by a Taylor approximate optimization paradigm. In comprehensive evaluations on four benchmark datasets, VMAN achieves superior performances under the (Generalized) TZSL setting.

Traditional methods treat the search problem as a process of selecting and ranking sequential documents. The methods have been proved effective and are widely used in the web search domain. However, due to the complexity and particularity of microblog text contents, the classical methods are rarely used microblog searches for specific topics. Focusing on the issue of searching for specific topics in microblog content, we present a microblog search method for security topics based on deep reinforcement learning by modeling the microblog search for specific topics as a continuous-state Markov decision process. We also design a novel deep Q network to evaluate the relevance of microblog content based on the target topic. We adopt reinforcement learning to solve the microblog search problem using an intelligent strategy and evaluate content relevance through deep learning. Experiments conducted on a real-world dataset show that our approach outperforms the selected baseline methods.

A facile oxalate co-precipitation method combined with solid-state sintering is used to synthesize layered LiNi0.6Co0.2Mn0.2O2 (NCM-622) as a low-cost cathode material for lithium ion batteries. Homogeneous Ni-Co-Mn oxalate precursors were obtained by adding oxalic acid into mixed metal acetates without the need of rigorous control of co-precipitation parameters. The effects of sintering temperature on the morphology, structure and electrochemical properties of the synthesized NCM-622 are studied. An intermediate sintering temperature of 850 °C is found to favor the integrity of the layered structure by compromising the solid-state reaction and Li loss, leading to a low degree of cation mixing. The as-synthesized NCM-622 delivers a capacity of 152 mA h g−1 after 100 cycles at 1 C (93.2% retention), and shows excellent rate capability with 76.8 mA h g−1 at 10 C, which is competitive to the materials prepared by the mainstream hydroxide co-precipitation method. Moreover, a solution-based surface modification was employed to further improve the cathode through self-adsorption of Al(OH)3 using Al2(SO4)3 precursor in order to form a Al2O3 coating. The facial surface modification has significantly improved the performance of NCM-622 under normal and high voltage conditions.

In recent years, deep neural networks have achieved remarkable successes in many pattern recognition tasks. However, the high computational cost and large memory overhead hinder them from applications on resource-limited devices. To address this problem, many deep network acceleration and compression methods have been proposed. One group of methods adopt decomposition and pruning techniques to accelerate and compress a pre-trained model. Another group designs single compact unit to stack their own networks. These methods are subject to complicated training processes, or lack of generality and extensibility. In this paper, we propose a general framework of architecture distillation, namely LightweightNet, to accelerate and compress convolutional neural networks. Rather than compressing a pre-trained model, we directly construct the lightweight network based on a baseline network architecture. The LightweightNet, designed based on a comprehensive analysis of the network architecture, consists of network parameter compression, network structure acceleration, and non-tensor layer improvement. Specifically, we propose the strategy of low-dimensional features of fully-connected layers for substantial memory saving, and design multiple efficient compact blocks to distill convolutional layers of baseline network with accuracy-sensitive distillation rule for notable time saving. Finally, it can effectively reduce the computational cost and the model size by > 4 × with negligible accuracy loss. Benchmarks on MNIST, CIFAR-10, ImageNet and HCCR (handwritten Chinese character recognition) datasets demonstrate the advantages of the proposed framework in terms of speed, performance, storage and training process. In HCCR, our method even outperforms traditional handcrafted features-based classifiers in terms of speed and storage while maintaining state-of-the-art recognition performance.

Lu(OH)CO3:Ln3+/LuBO3:Ln3+ (Ln = Eu, Tb) nanospheres have been synthesized by the soft chemistry route. This synthesis process does not require the addition of any organic solvents and surfactant. The whole process consists of homogeneous precipitation and hydrothermal process. First, uniform spherical Lu(OH)CO3:Ln3+ (Ln = Eu, Tb) nanoparticles with the sizes of 70 nm were obtained by the classic urea-based homogeneous precipitation. Then, with the above Lu(OH)CO3:Ln3+ (Ln = Eu, Tb) nanosphere as the precursor, the composites of Lu(OH)CO3:Ln3+/LuBO3:Ln3+ (Ln = Eu, Tb) nanospheres formed by the simple hydrothermal process. The phases, morphologies, as well as the luminescence properties of as-prepared samples were characterized by means of X-ray diffraction, field-emission scanning electron microscopy, transmission electron microscopy, and photoluminescence. Under a short wavelength excitation, the down-conversion Eu3+ and Tb3+-doped LuBO3 samples were investigated and explained.

Smoldering combustion frequently occurs in the mined-out gobs of coal mines. This can act as an ignition source, triggering gas explosions and resulting in serious accidents. In this study, a smoldering test device was constructed, in which horizontal, vertical, forward and reverse smoldering combustion experiments were performed to explore the smoldering characteristics (dynamic development process, smoldering rate and temperature, etc.) of a coal pile. It is found that owing to the evaporation moisture present in the coal and its re-condensation on the coal ahead of the smoldering front, a constant temperature of approximately 70–90 °C is maintained during the spread of forward smoldering. In the initial stage of reverse smoldering, both the pyrolysis and oxidation fronts exist and propagate backward; meanwhile, in the later stage, the intensity of the endothermic pyrolysis reaction is greatly reduced, resulting that the maximum smoldering temperature of reverse smoldering is slightly higher than that of the forward smoldering. The vertical smoldering temperature is 1.5–2 times that of horizontal smoldering, owing to the poorer oxygen supply in the horizontal spread. The propagation of the oxidation front is extremely slow, and is in the order of 0.1 mm/min; this makes it difficult to characterize and detect the smoldering fire, especially in inaccessible mined-out gobs with complex environments. This study provides useful insights into the identification of smoldering fire and assesses its development stage.

In view of the potential application of NIR luminescence in biology, a series of four new 3,5-Heptanedione (β-diketonate) ytterbium complexes, e.g. [Yb(L)3(H2O)2] (1), [Yb(L)3(2,2′-bpy)] (2), [Yb(L)3(dpq)] (3) and [Yb(L)3(dppz)] (4) have been designed and isolated. Crystal structures analysis show that all complexes 1–4 exhibit the similar mononuclear structure, in which the central Yb3+ ion is eight-coordinated forming the square-anti-prismatic geometry. NIR luminescence of all complexes 1–4 shows strong characteristic emission of Yb3+ ions. To improve the mechanical and thermal stability, complex 1 was further doped into poly-(methyl methacrylate) (PMMA) for preparing a thin film (1@PMMA). Their thermal stability and NIR luminescence were essentially enhanced in contrast to that for complex 1. This NIR luminescent thin film may be of potential applications in biological dictation and luminescent probe.

In this study, bamboo flour (BF) was chemically modified using alkyl ketene dimer (AKD) or alkali treatment for the preparation of high-density polyethylene/bamboo flour (HDPE/BF) composites with or without compatibilizer maleic anhydride grafted high-density polyethylene (HDPE-g-MAH). The modified BF was characterized by Fourier transformation infrared (FTIR) spectroscopy and contact angle measurement, and the mechanical properties, morphology and water resistance of the HDPE/BF composites were investigated and compared. The results show that highly hydrophobic BF can be obtained by a simple procedure using a common paper sizing agent AKD. AKD can react with the hydroxyl groups in cellulose to form β-keto ester. The combination of AKD treatment and addition of HDPE-g-MAH is more effective in improving the mechanical properties of HDPE/BF composites than using them alone. The impact strength, flexural strength and modulus of HDPE/A-BF/MAH composite are increased by 100.0%, 197.1% and 61.8% compared with that of HDPE/U-BF composites, and the water absorption and thickness swelling are decreased by 87.7% and 87.4% after immersion in water for one week, respectively. These results indicate that HDPE/BF composites with good mechanical properties and high water resistance can be obtained by AKD treatment.

The vast development of nanofabrication has spurred recent progress for the manipulation of light down to a region much smaller than the wavelength. Metallic plasmonic array structures are demonstrated to be the most powerful platform to realize controllable light-matter interactions and have found wide applications due to their rich and tunable optical performance through the morphology and parameter engineering. Here, various light-management mechanisms that may exist on metallic plasmonic array structures are described. Then, the typical techniques for fabrication of metallic plasmonic arrays are summarized. Next, some recent applications of plasmonic arrays are reviewed, including plasmonic sensing, surface-enhanced spectroscopies, plasmonic nanolasing, and perfect light absorption. Lastly, the existing challenges and perspectives for metallic plasmonic arrays are discussed. The aim is to provide guidance for future development of metallic plasmonic array structures.

Deep learning technology has been extensively explored by existing methods to improve the performance of target detection in remote sensing images, due to its powerful feature extraction and representation abilities. However, these methods usually focus on the interior features of the target, but ignore the exterior semantic information around the target, especially the object-level relationship. Consequently, these methods fail to detect and recognize targets in the complex background where multiple objects crowd together. To handle this problem, a diversified context information fusion framework based on convolutional neural network (DCIFF-CNN) is proposed in this paper, which employs the structured object-level relationship to improve the target detection and recognition in complex backgrounds. The DCIFF-CNN is composed of two successive sub-networks, i.e., a multi-scale local context region proposal network (MLC-RPN) and an object-level relationship context target detection network (ORC-TDN). The MLC-RPN relies on the fine-grained details of objects to generate candidate regions in the remote sensing image. Then, the ORC-TDN utilizes the spatial context information of objects to detect and recognize targets by integrating an attentional message integrated module (AMIM) and an object relational structured graph (ORSG). The AMIM is integrated into the feed-forward CNN to highlight the useful object-level context information, while the ORSG builds the relations between a set of objects by processing their appearance features and geometric features. Finally, the target detection method based on DCIFF-CNN effectively represents the interior and exterior information of the target by exploiting both the multiscale local context information and the object-level relationships. Extensive experiments are conducted, and experimental results demonstrate that the proposed DCIFF-CNN method improves the target detection and recognition accuracy in complex backgrounds, showing superiority to other state-of-the-art methods.

Scene text detection has been studied extensively. Existing methods detect either words or text lines and use either word-level or line-level annotated data for training. In this paper, we propose a dual-task network that can perform word-level and line-level text detection simultaneously and use training data of both levels of annotation to boost the performance. The dual-task network has two detection heads for word-level and line-level text detection, respectively. Then we propose a mutual guidance scheme for the joint training of the two tasks with two modules: line filtering module utilizes the output feature map of the text line detector to filter out the non-text regions for the word detector, and word enhancing module provides prior positions of words for the text line detector depending on the output feature map of the word detector. Experimental results of word-level and line-level text detection demonstrate the effectiveness of the proposed dual-task network and mutual guidance scheme, and the results of our method are competitive with state-of-the-art methods.

This paper considers recognizing texts shown in a source language and translating into a target language, without generating the intermediate source language text image recognition results. We call this problem Cross-Lingual Text Image Recognition (CLTIR). To solve this problem, we propose a multi-task system containing a main task of CLTIR and an auxiliary task of Mono-Lingual Text Image Recognition (MLTIR) simultaneously. Two different sequence to sequence learning methods, a convolution based attention model and a Bidirectional Long Short-Term Memory (BLSTM) model with Connectionist Temporal Classification (CTC), are adopted for these tasks respectively. We evaluate the system on a newly collected Chinese-English bilingual movie subtitle image dataset. Experimental results demonstrate the multi-task learning framework performs superiorly in both languages.

Recent research has revealed that deep neural networks are highly vulnerable to adversarial examples. In this paper, different from most adversarial attacks which directly modify pixels in spatial domain, we propose a novel black-box attack in frequency domain, named as f-mixup, based on the property of natural images and perception disparity between human-visual system (HVS) and convolutional neural networks (CNNs): First, natural images tend to have the bulk of their Fourier spectrums concentrated on the low frequency domain; Second, HVS is much less sensitive to high frequencies while CNNs can utilize both low and high frequency information to make predictions. Extensive experiments are conducted and show that deeper CNNs tend to concentrate more on the higher frequency domain, which may explain the contradiction between robustness and accuracy. In addition, we compared f-mixup with existing attack methods and observed that our approach possesses great advantages. Finally, we show that f-mixup can be also incorporated in training to make deep CNNs defensible against a kind of perturbations effectively.

Surface-enhanced Raman spectroscopy (SERS) is a highly sensitive technique with molecular specificity, making it an ideal analytical tool in various fields. However, the breadth of practical applications of SERS has been severely limited because it is still a great challenge to achieve simultaneously a high sensitivity and a high reproducibility. Herein, we report a highly controllable method to fabricate periodic bowtie SERS substrates with a narrow nanogap, high SERS enhancement, and good uniformity over a large area. The periodic bowtie template is first fabricated over a gold film by holographic lithography (HL), followed by Au deposition to obtain a conductive plasmonic bowtie array. The gap size is then narrowed down by pulsed electrodeposition of Ag simultaneously monitored in situ by electrochemical dark field spectroscopy. Thus, we are able to observe the most sensitive change in the scattering spectra when the gap is just about to merge and obtain uniform SERS substrates with a gap size down to around 5 nm. The average enhancement factor of 5 × 107 to 1 × 108 is obtained, which is 50 times larger than that from Au nanoparticle-assembled substrates and 140 times larger than that from commercial Klarite chips. This substrate offers a promising opportunity for SERS practical applications.

Face age editing has become a crucial task in film post-production, and is also becoming popular for general purpose photography. Recently, adversarial training has produced some of the most visually impressive results for image manipulation, including the face aging/de-aging task. In spite of considerable progress, current methods often present visual artifacts and can only deal with low-resolution images. In order to achieve aging/de-aging with the high quality and robustness necessary for wider use, these problems need to be addressed. This is the goal of the present work. We present an encoder-decoder architecture for face age editing. The core idea of our network is to encode a face image to age-invariant features, and learn a modulation vector corresponding to a target age. We then combine these two elements to produce a realistic image of the person with the desired target age. Our architecture is greatly simplified with respect to other approaches, and allows for fine-grained age editing on high resolution images in a single unified model. Source codes are available at https://github.com/InterDigitalInc/HRFAE.

Computational analysis of human multimodal sentiment is an emerging research area. Fusing semantic, visual and acoustic modalities requires exploring the inter-modal and intra-modal interactions. The first challenge for the inter-modal understanding is to break the heterogeneous gap between different modalities. Meanwhile, when modeling the intra-modal connection of time-series data, we must deal with the long-range dependencies among multiple steps. Moreover, The time-series data usually is unaligned between different modalities because individually specialized processing approaches or sampling frequencies. In this paper, we propose a method based on the transformer and the gate mechanism - the Gate-Fusion Transformer - to address these problems. We conducted detailed experiments for verifying the effectiveness of our proposed method. Because of the flexibility of gate-mechanism for information flow controlling and the great modeling power of the transformer for modeling the inter- and intra-modal interactions, we can achieve superior performance compared with the current state-of-the-art method but more extendible and flexible by stacking multiple gate-fusion blocks.

As camera-based documents are increasingly used, the rectification of distorted document images becomes a need to improve the recognition performance. In this paper, we propose a novel framework for both rectifying distorted document image and removing background finely, by estimating pixel-wise displacements using a fully convolutional network (FCN). The document image is rectified by transformation according to the displacements of pixels. The FCN is trained by regressing displacements of synthesized distorted documents, and to control the smoothness of displacements, we propose a Local Smooth Constraint (LSC) in regularization. Our approach is easy to implement and consumes moderate computing resource. Experiments proved that our approach can dewarp document images effectively under various geometric distortions, and has achieved the state-of-the-art performance in terms of local details and overall effect.

Feature representation learning, an emerging topic in recent years, has achieved great progress. Powerful learned features can lead to excellent classification accuracy. In this article, a selective and robust feature representation framework with a supervised constraint (SRSC) is presented. SRSC seeks a selective, robust, and discriminative subspace by transforming the original feature space into the category space. Particularly, we add a selective constraint to the transformation matrix (or classifier parameter) that can select discriminative dimensions of the input samples. Moreover, a supervised regularization is tailored to further enhance the discriminability of the subspace. To relax the hard zero-one label matrix in the category space, an additional error term is also incorporated into the framework, which can lead to a more robust transformation matrix. SRSC is formulated as a constrained least square learning (feature transforming) problem. For the SRSC problem, an inexact augmented Lagrange multiplier method (ALM) is utilized to solve it. Extensive experiments on several benchmark data sets adequately demonstrate the effectiveness and superiority of the proposed method. The proposed SRSC approach has achieved better performances than the compared counterpart methods.

Most existing text spotting methods either focus on horizontal/oriented texts or perform arbitrary shaped text spotting with character-level annotations. In this paper, we propose a novel text spotting framework to detect and recognize text of arbitrary shapes in an end-to-end manner, using only word/line-level annotations for training. Motivated from the name of TextSnake, which is only a detection model, we call the proposed text spotting framework TextDragon. In TextDragon, a text detector is designed to describe the shape of text with a series of quadrangles, which can handle text of arbitrary shapes. To extract arbitrary text regions from feature maps, we propose a new differentiable operator named RoISlide, which is the key to connect arbitrary shaped text detection and recognition. Based on the extracted features through RoISlide, a CNN and CTC based text recognizer is introduced to make the framework free from labeling the location of characters. The proposed method achieves state-of-the-art performance on two curved text benchmarks CTW1500 and Total-Text, and competitive results on the ICDAR 2015 Dataset.

A series of isomorphic lanthanide metal–organic frameworks (Ln-MOFs), namely [Ln(dcbba)(DMF)2]n·H2O·0.5DMF [Ln = La (1), Eu (2) and Tb (3); H3dcbba = 4-(3, 5-dicarboxylatobenzyloxy)benzoic acid; DMF = N,N′-dimethylformamide] have been designed and synthesized by the solvothermal reactions of H3dcbba and La(NO3)3·6H2O. Single-crystal X-ray diffraction analysis reveals that the complexes 1–3 exhibit a (3,6)-connected open framework structure with binuclear [Ln2(COO)6]n secondary building units as 6-connected nodes and H3dcbba ligands as 3-connected nodes. The isostructural mixed La/Eu/Tb-dcbba (4) was obtained via the in-suit doping of different Ln3+ ions into the host framework, which is able to emit pure orange, white, and blue light when excited at 300, 305, and 350 nm, respectively. Subsequently, a novel and multifunctional sensing process was designed based on the excitation wavelength sensitive color tunable luminescent sample 4, which can detect HS– ions, THF (tetrahydrofuran), and Ag+ ions via diffe...

Polypyrrole (PPy) is a highly conductive polymer and promising pseudocapacitive material for energy storage. However, the poor stability of PPy based supercapacitors (SCs) limits their applications. Here we demonstrate a facile dipping process to uniformly grow 4-butylbenzenesulfonate modified PPy film on a filter paper substrate. The flexible and free-standing PPy paper electrode demonstrated excellent cycling stability with capacitance retention of ~84.8% after 200,000 cycles. We believe that the 4-butylbenzenesulfonate modification could increase the order degree of PPy via layer-by-layer stacked α-α coupled chains, leading to a robust structure under the charging and discharging progress. This work pushes the cycling limitation of PPy based SCs and realizes their potentials for practical applications.

In this paper, we propose two kinds of feature extracting frameworks that can extract cascaded class-specific and class-mixture features, respectively, by taking the restricted Boltzmann machine (RBM) as the basic building blocks; we further call them as a CS-RBM and CM-RBM feature extractor. The discriminations of features from both CS-RBM and CM-RBM are verified better than the class-independent (traditional) RBM (CI-RBM) feature extractor. As one mini-batch samples are randomly selected from all classes during the training phase of the traditional RBM, which can make that the above mini-batch data contain easy-confusing samples from different categories. Therefore, the features from CI-RBM are difficult to distinguish these samples from the confused categories. CS-RBM and CM-RBM can overcome the above sample confusing problem efficiently and effectively. To cope with the real-valued input samples, we further extend the binary RBM to Gaussian–Bernoulli RBM (GBRBM), leading to the CS-GBRBM (CM-GBRBM) feature extracting framework. Experiments on binary datasets, i.e., MNIST and USPS, scene image dataset (Scene-15), and object image dataset (Coil-100), well verify the above facts and show the competitive results.

Tip-enhanced Raman spectroscopy (TERS), known as nanospectroscopy, has received increasing interest as it can provide nanometer spatial resolution and chemical fingerprint information of samples simultaneously. Since Ag tips are well accepted to show a higher TERS enhancement than that of gold tips, there is an urgent quest for Ag TERS tips with a high enhancement, long lifetime, and high reproducibility, especially for atomic force microscopy (AFM)-based TERS. Herein, we developed an electrodeposition method to fabricate Ag-coated AFM TERS tips in a highly controllable and reproducible way. We investigated the influence of the electrodeposition potential and time on the morphology and radius of the tip. The radii of Ag-coated AFM tips can be rationally controlled at a few to hundreds nanometers, which allows us to systematically study the dependence of the TERS enhancement on the tip radius. The Ag-coated AFM tips show the highest TERS enhancement under 632.8 nm laser excitation and a broad localized surface plasmon resonance (LSPR) response when coupled to a Au substrate. The tips exhibit a lifetime of 13 days, which is particularly important for applications that need a long measuring time.

The identically and independently distributed (i.i.d.) condition required by conventional machine learning approaches may sometimes be violated when patterns occur as groups (where each group shares a homogeneous style, called a field). By breaking it, we extend in this paper the famous Support Vector Machine (SVM) to a novel framework named Field Support Vector Machine (F-SVM), in which the training and predicting a group of patterns (i.e., a field pattern) are performed simultaneously. Specifically, the proposed F-SVM is learned by optimizing simultaneously both the classifier and the Style Normalization Transformation (SNT) for each group of data, even feasible in the high-dimensional kernel space. The SNT transform the original style-discriminative patterns to style-free ones, satisfying the i.i.d. assumption required by the conventional SVM learning and implementation. An efficient optimization algorithm is further developed with the convergence guaranteed theoretically. More importantly, by appropriately exploring the style consistency in each field, the proposed F-SVM model is able to significantly improve the classification accuracy. A series of experiments are conducted to verify the effectiveness and confirmed improvement on the performance of the F-SVM model. Empirical results show that the proposed F-SVM outperforms other relevant baselines in two different benchmark data sets.

Fine-grained recognition is one of the most difficult topics in visual recognition, which aims at distinguishing confusing categories such as bird species within a genus. The information of part and bounding boxes in fine-grained images is very important for improving the performance. However, in real applications, the part and/or bounding box annotations may not exist. This makes fine-grained recognition a challenging problem. In this paper, we propose a jointly trained Convolutional Neural Network (CNN) architecture to solve the fine-grained recognition problem without using part and bounding box information. In this framework, we first detect part candidates by calculating the gradients of feature maps of a trained CNN model w.r.t. the input image and then filter out unnecessary ones by fusing two saliency detection methods. Meanwhile, two groups of global object locations are obtained based on the saliency detection methods and a segmentation method. With the filtered part candidates and approximate object locations as inputs, we construct the CNN architecture with local parts and global discrimination (LG-CNN) which consists of two CNN networks with shared weights. The upper stream of LG-CNN is focused on the part information of the input image, the bottom stream of LG-CNN is focused on the global input image. LG-CNN is jointly trained by two stream loss functions to guide the updating of the shared weights. Experiments on three popular fine-grained datasets well validate the effectiveness of our proposed LG-CNN architecture. Applying our LG-CNN architecture to generic object recognition datasets also yields superior performance over the directly fine-tuned CNN architecture with a large margin.

Expansion of mudstone through absorbing water can increase local stress in information, which is one of the main reasons for casing damage in mudstone. To establish a reasonable mechanical model of local stress increase is the key problem for casing damage in locally expansive mudstone. Many models of uniform stress or linear stress are established to simulate the local stress increase. However, the accuracy and scientificity for those models are insufficient. In this paper, a study on casing strength in locally expansive mudstone is carried out based on theoretical analysis and numerical calculation. Firstly, a mechanical model of casings under parabolic-like loads is established to simulate the local stress increase. Then, an analytical solution of casings is obtained based on the stress function method. Next, a finite element calculation of casings under the same boundary condition is carried out. In the end, the comparisons between analytical solution and finite element method (FEM) are conducted. The results reveal that the theoretical solutions are in good agreement with the FEM. In addition, a novel predicting method of the casing stress in locally expansive mudstone is presented. Finally, our theoretical analysis can be the standard for numerical and approximate calculations. It is helpful to explore the casing damage mechanism in mudstone layer and reduce the casing collapse in complex underground environment.

In this work, commercial LiCoO 2 is modified with a glassy B 2 O 3 by solution mixing with H 3 BO 3 followed by post-calcination in order to enhance its high-voltage electrochemical performance. The glassy B 2 O 3 coating/additive is believed to serve as an effective physiochemical buffer and protection between LiCoO 2 and liquid electrolyte, which can suppress the high-voltage induced electrolyte decomposition and active material dissolution. During the early cycling and due to the electrochemical force, the as-coated B 2 O 3 glasses which have 3D open frameworks tend to accommodate some mobile Li + and form a more chemically-resistant and ion-conductive lithium boron oxide (LBO) interphase as a major component of the solid electrolyte interphase (SEI), which consequently enables much easier Li + diffusion/transfer at the solid-liquid interfaces upon further cycling. Due to the synergetic effects of B 2 O 3 coating/modification, the high-voltage capacity and energy density of the B 2 O 3 -modified LiCoO 2 cathode are promisingly improved by 35% and 30% after 100 cycles at 1 C within 3.0–4.5 V vs. Li/Li + . Meanwhile, the high-rate performance of the B 2 O 3 -modified electrode is even more greatly improved, showing a capacity of 105 mAh g −1 at 10 C while the bare electrode has dropped to no more than 30 mAh g −1 under this rate condition.

An adsorption–coagulation–microfiltration (A–C–MF) hybrid process using copper ferrocyanide (CuFC) as the adsorbent was developed in this study for the removal of cesium from water. The effect of coagulant on cesium removal by CuFC adsorption was investigated. A simpler calculation method for determining the CuFC dosage was established based on the Freundlich adsorption isotherms. Although the volume of treated water was 1.32 times higher for the A–C–MF process than the adsorption–microfiltration (A–MF) process, the cake layer and membrane pore resistances in the former were both lower than those in the latter due to the coagulation process.

Writer identification is an important topic for pattern recognition and artificial intelligence. Traditional methods rely heavily on sophisticated hand-crafted features to represent the characteristics of different writers. In this paper, we propose an end-to-end framework for online text-independent writer identification by using a recurrent neural network (RNN). Specifically, the handwriting data of a particular writer are represented by a set of random hybrid strokes (RHSs). Each RHS is a randomly sampled short sequence representing pen tip movements ( $xy$ -coordinates) and pen-down or pen-up states. RHS is independent of the content and language involved in handwriting; therefore, writer identification at the RHS level is more general and convenient than the character level or the word level, which also requires character/word segmentation. The RNN model with bidirectional long short-term memory is used to encode each RHS into a fixed-length vector for final classification. All the RHSs of a writer are classified independently, and then, the posterior probabilities are averaged to make the final decision. The proposed framework is end-to-end and does not require any domain knowledge for handwriting data analysis. Experiments on both English (133 writers) and Chinese (186 writers) databases verify the advantages of our method compared with other state-of-the-art approaches.

Bag of Words (BoW) model and Convolutional Neural Network (CNN) are two milestones in visual recognition. Both BoW and CNN require a feature pooling operation for constructing the frameworks. Particularly, the max-pooling has been validated as an efficient and effective pooling method compared with other methods such as average pooling and stochastic pooling. In this paper, we first evaluate different pooling methods, and then propose a new feature pooling method termed as selective, discriminative and equalizing pooling (SDE). The SDE representation is a feature learning mechanism by jointly optimizing the pooled representations with the target of learning more selective, discriminative and equalizing features. We use bilevel optimization to solve the joint optimization problem. Experiments on seven benchmark datasets (including both single-label and multi-label ones) well validate the effectiveness of our framework. Particularly, we achieve the state-of-the-art fused results (mAP) of 93.21 and 93.97% on the PASCAL VOC2007 and VOC2012 datasets, respectively.

With the increasing demand for high energy density in portable-device Li-ion batteries (LIBs), efforts are devoted to increase and stabilize the capacity of LiCoO2 at high operation voltages. Herein, we report a low-cost and eco-friendly wet-chemical method to coat Al2O3 on LiCoO2, using only aluminium sulphate and water as source materials. A nanoscale oxide layer is coated on the surface of LiCoO2 particles through hydrogen-bonding assisted adsorption of the hydrolysed Al(OH)3 nanoparticles. The as-proposed Al2O3-coating provides excellent physico-chemical protection and kinetically-favourable interfaces for the LiCoO2 electrode, resulting in remarkable improvements of the electrode's cycling stability and rate capability when tested at high cutoff voltages up to 4.7 V (vs. Li/Li+). The synergetic effects of the oxide coating, e.g. alleviated electrolyte decomposition and reduced generation of irreversible solid electrolyte interphase (SEI) constituents (LiF/Li2CO3 and organics), are attributed to the improvements. At the cutoff voltage of 4.5 V, the modified LiCoO2 electrode in this work exhibits excellent cycling stability (147 mA h g−1, 82.6% retention after 500 cycles at 1C) and competitive rate capability (130 mA h g−1 at 10C), which are some of the best results reported so far. The outstanding high-voltage electrode performance and the simple and scalable coating approach show great promise of LiCoO2 cathodes in future high-energy and high-power LIBs.

Recent deep learning based methods have achieved the state-of-the-art performance for handwritten Chinese character recognition (HCCR) by learning discriminative representations directly from raw data. Nevertheless, we believe that the long-and-well investigated domain-specific knowledge should still help to boost the performance of HCCR. By integrating the traditional normalization-cooperated direction-decomposed feature map (directMap) with the deep convolutional neural network (convNet), we are able to obtain new highest accuracies for both online and offline HCCR on the ICDAR-2013 competition database. With this new framework, we can eliminate the needs for data augmentation and model ensemble, which are widely used in other systems to achieve their best results. This makes our framework to be efficient and effective for both training and testing. Furthermore, although directMap+convNet can achieve the best results and surpass human-level performance, we show that writer adaptation in this case is still effective. A new adaptation layer is proposed to reduce the mismatch between training and test data on a particular source layer. The adaptation process can be efficiently and effectively implemented in an unsupervised manner. By adding the adaptation layer into the pre-trained convNet, it can adapt to the new handwriting styles of particular writers, and the recognition accuracy can be further improved consistently and significantly. This paper gives an overview and comparison of recent deep learning based approaches for HCCR, and also sets new benchmarks for both online and offline HCCR.

A white-light-emitting LnCP is effectively applied to the detection of Ag+ and Mn2+ ions with high selectivity and sensitivity via luminescence-colour change.

The winter is very cold in cold and severe cold areas in China, we will burn coal to keep warm, in order to allow urban and rural residents to be warm in winter. Uncontrolled resource development and inefficient energy utilization have resulted in a lot of waste of resources and increasingly serious ecological pollution. How to develop and utilize renewable resources is a serious challenge that China needs to face. This paper carries out a study on the solar-natural gas combined heating system based on the climate and energy situation of cold and severe cold regions, to achieve a seamless connection between the two systems. This technology can combine the advantages of two energy supply systems, overcome their own shortcomings. It has obvious energy-saving effect, significant economic benefits and a high promotional value.

How to efficiently judge whether a natural image contains texts or not is an important problem. Since text detection and recognition algorithms are usually time-consuming, and it is unnecessary to run them on images that do not contain any texts. In this paper, we investigate this problem from two perspectives: the speed and the accuracy. First, to achieve high speed for efficient filtering large number of images especially on CPU, we propose using small and shallow convolutional neural network, where the features from different layers are adaptively pooled into certain sizes to overcome difficulties caused by multiple scales and various locations. Although this can achieve high speed but its accuracy is not satisfactory due to limited capacity of small network. Therefore, our second contribution is using the knowledge distillation to improve the accuracy of the small network, by constructing a larger and deeper neural network as teacher network to instruct the learning process of the small network. With the above two strategies, we can achieve both high speed and high accuracy for filtering scene text images. Experimental results on a benchmark dataset have shown the effectiveness of our method: the teacher network yields state-of-the-art performance, and the distilled small network achieves high performance while maintaining high speed which is 176 times faster on CPU and 3.8 times faster on GPU than a compared benchmark method.

In contrast to machine recognizers that rely on training with large handwriting data, humans can recognize handwriting accurately on learning from few samples, and can even generalize to handwritten characters from printed samples. Simulating this ability in machine recognition is important to alleviate the burden of labeling large handwriting data, especially for large category set as in Chinese text. In this paper, inspired by human learning, we propose a cross-modal prototype learning (CMPL) method for zero-shot online handwritten character recognition: for unseen categories, handwritten characters can be recognized without learning from handwritten samples, but instead from printed characters. Particularly, the printed characters (one for each class) are embedded into a convolutional neural network (CNN) feature space to obtain prototypes representing each class, while the online handwriting trajectories are embedded with a recurrent neural network (RNN). Via cross-modal joint learning, handwritten characters can be recognized according to the printed prototypes. For unseen categories, handwritten characters can be recognized by only feeding a printed sample per category. Experiments on a benchmark Chinese handwriting database have shown the effectiveness and potential of the proposed method for zero-shot handwriting recognition.

This paper introduces a Chinese Ancient Handwritten Characters Database (CASIA-AHCDB) for character recognition research. The database was built by annotating 11,937 pages of Chinese ancient handwritten documents. It consists of more than 2.2 million annotated handwritten character samples of 10,350 categories. According to the source of these documents, the database is divided into two datasets of different styles: Complete Library in Four Sections (AHCDB-style1) and Ancient Buddhist Scriptures (AHCDB-style2). Each dataset can be divided into three parts based on its applications. The first part, called basic category set, contains samples of common categories in two datasets, and is suitable for basic character recognition task. The second part, called enhanced category set, is mainly used for open-set character recognition task based on the basic character recognition. The third part, called the reserved category set, can be used in many pattern recognition tasks in the future. Based on the large category set, the various writing styles and the imbalanced sample number per category, CASIA-AHCDB can also be used for various classification and learning tasks such as transfer learning, few-shot learning. We performed experiments of basic character recognition on the basic category set, and report the results for benchmark. More techniques can be evaluated on this challenging database in the future.

A simple imine based receptor NA-1 has been synthesized for detection of Zinc ions. Probe NA-1 showed the selective colorimetric changes with Zinc (II) ions whereas other metal ions didn't showed any observable colorimetric changes. The probe showed the very selective turn-on fluorescence response with Zn(II) ions among other rival metal ions like Cd(II) and Hg(II). The mode of binding was studied by 1H-nmr titrations and fluorescence spectroscopy. Jobs plot analysis confirming that the probe NA-1 forms 1:1 complex with Zn(II). The observed fluorescence and absorption change further supported by theoretical calculations. The turn-on fluorescence of the probe NA-1 is probably attributable to the interruption of intramolecular charge transfer as well as ESIPT. The limit of detection of the probe for Zn(II) sensing is in the range of 14 nano molar. Cytotoxicity (MTT) assay of the probe in live HeLa cells is showing that the probe is least toxic to cells. The probe NA-1 is effectively applied to detect Zn(II) ions in HeLa cells and suggesting the probe is NA-l permeable to cell wall and viable for Zinc(II) ions imaging in live cells.

Motivated by the fact that humans can grasp semantic-invariant features shared by the same category while attention-based models focus mainly on discriminative features of each object, we propose a scalable paired adversarial learning (PAL) method for image-to-markup generation. PAL can incorporate the prior knowledge of standard templates to guide the attention-based model for discovering semantic-invariant features when the model pays attention to regions of interest. Furthermore, we also extend the convolutional attention mechanism to speed up the image-to-markup parsing process while achieving competitive performance compared with recurrent attention models. We evaluate the proposed method in the scenario of handwritten-image-to-LaTeX generation, i.e., converting handwritten mathematical expressions to LaTeX. Experimental results show that our method can significantly improve the generalization performance over standard attention-based encoder-decoder models.

As one of the major factors of the nontechnical losses (NTLs) in distribution networks, the electricity theft causes significant harm to power grids, which influences power supply quality and reduces operating profits. In order to help utility companies solve the problems of inefficient electricity inspection and irregular power consumption, a novel hybrid convolutional neural network-random forest (CNN-RF) model for automatic electricity theft detection is presented in this paper. In this model, a convolutional neural network (CNN) firstly is designed to learn the features between different hours of the day and different days from massive and varying smart meter data by the operations of convolution and downsampling. In addition, a dropout layer is added to retard the risk of overfitting, and the backpropagation algorithm is applied to update network parameters in the training phase. And then, the random forest (RF) is trained based on the obtained features to detect whether the consumer steals electricity. To build the RF in the hybrid model, the grid search algorithm is adopted to determine optimal parameters. Finally, experiments are conducted based on real energy consumption data, and the results show that the proposed detection model outperforms other methods in terms of accuracy and efficiency.