arXiv subjects

Jing Zhang

Publications and source records attributed to Jing Zhang.

At least 847 records · Page 47Linked to original sources

arXiv2019

Metal oxide nanomaterials for pseudocapacitors

With the rapid development of economy, the consumption of fossil fuels, and the increasing environment pollution, there is in urgent need of seeking clean and renewable energy sources, as well as highly efficient and low-cost energy storage technologies. Concerning electrochemically active materials, transition metal oxides (TMOs) are regarded as the most promising candidates for the next generation SCs and have been widely reported. Therefore, in this review, we mainly pay attention to TMOs with excellent electrochemical performance. This review is divided into four parts: (a) the background including energy storage mechanisms and electrochemical characterizations of SCs; (b) various kinds of TMOs for SCs are analyzed in detail involved in crystal structure, conductivity, and energy storage mechanism; (c) hybridizations with other materials to improve the electrochemical performance of TMOs are summarized and discussed; (d) a perspective and short remarks about the development and design of TMOs for superior performance SCs are shown.

physics.app-ph↗

arXiv2019

Deep Fault Diagnosis for Rotating Machinery with Scarce Labeled Samples

Early and accurately detecting faults in rotating machinery is crucial for operation safety of the modern manufacturing system. In this paper, we proposed a novel Deep fault diagnosis (DFD) method for rotating machinery with scarce labeled samples. DFD tackles the challenging problem by transferring knowledge from shallow models, which is based on the idea that shallow models trained with different hand-crafted features can reveal the latent prior knowledge and diagnostic expertise and have good generalization ability even with scarce labeled samples. DFD can be divided into three phases. First, a spectrogram of the raw vibration signal is calculated by applying a Short-time Fourier transform (STFT). From those spectrograms, discriminative time-frequency domain features can be extracted and used to form a feature pool. Then, several candidate Support vector machine (SVM) models are trained with different combinations of features in the feature pool with scarce labeled samples. By evaluating the pretrained SVM models on the validation set, the most discriminative features and best-performed SVM models can be selected, which are used to make predictions on the unlabeled samples. The predicted labels reserve the expert knowledge originally carried by the SVM model. They are combined together with the scarce fine labeled samples to form an Augmented training set (ATS). Finally, a novel 2D deep Convolutional neural network (CNN) model is trained on the ATS to learn more discriminative features and a better classifier. Experimental results on two fault diagnosis datasets demonstrate the effectiveness of the proposed DFD, which achieves better performance than SVM models and the vanilla deep CNN model trained on scarce labeled samples. Moreover, it is computationally efficient and is promising for real-time rotating machinery fault diagnosis.

eess.SP↗

arXiv2019

Semi-Supervised Graph Embedding for Multi-Label Graph Node Classification

The graph convolution network (GCN) is a widely-used facility to realize graph-based semi-supervised learning, which usually integrates node features and graph topologic information to build learning models. However, as for multi-label learning tasks, the supervision part of GCN simply minimizes the cross-entropy loss between the last layer outputs and the ground-truth label distribution, which tends to lose some useful information such as label correlations, so that prevents from obtaining high performance. In this paper, we pro-pose a novel GCN-based semi-supervised learning approach for multi-label classification, namely ML-GCN. ML-GCN first uses a GCN to embed the node features and graph topologic information. Then, it randomly generates a label matrix, where each row (i.e., label vector) represents a kind of labels. The dimension of the label vector is the same as that of the node vector before the last convolution operation of GCN. That is, all labels and nodes are embedded in a uniform vector space. Finally, during the ML-GCN model training, label vectors and node vectors are concatenated to serve as the inputs of the relaxed skip-gram model to detect the node-label correlation as well as the label-label correlation. Experimental results on several graph classification datasets show that the proposed ML-GCN outperforms four state-of-the-art methods.

cs.LG↗

arXiv2019

FAMED-Net: A Fast and Accurate Multi-scale End-to-end Dehazing Network

Single image dehazing is a critical image pre-processing step for subsequent high-level computer vision tasks. However, it remains challenging due to its ill-posed nature. Existing dehazing models tend to suffer from model overcomplexity and computational inefficiency or have limited representation capacity. To tackle these challenges, here we propose a fast and accurate multi-scale end-to-end dehazing network called FAMED-Net, which comprises encoders at three scales and a fusion module to efficiently and directly learn the haze-free image. Each encoder consists of cascaded and densely connected point-wise convolutional layers and pooling layers. Since no larger convolutional kernels are used and features are reused layer-by-layer, FAMED-Net is lightweight and computationally efficient. Thorough empirical studies on public synthetic datasets (including RESIDE) and real-world hazy images demonstrate the superiority of FAMED-Net over other representative state-of-the-art models with respect to model complexity, computational efficiency, restoration accuracy, and cross-set generalization. The code will be made publicly available.

cs.CV↗

arXiv2019

Rado's conjecture and its Baire version

Rado's Conjecture is a compactness/reflection principle that says any nonspecial tree of height $ω_1$ has a nonspecial subtree of size $\leq \aleph_1$. Though incompatible with Martin's Axiom, Rado's Conjecture turns out to have many interesting consequences that are consequences of forcing axioms. In this paper, we obtain consistency results concerning Rado's Conjecture and its Baire version. In particular, we show a fragment of PFA, that is the forcing axiom for \emph{Baire Indestructibly proper forcings}, is compatible with the Baire Rado's Conjecture. As a corollary, Baire Rado's Conjecture does not imply Rado's Conjecture. Then we discuss the strength and limitations of the Baire Rado's Conjecture regarding its interaction with simultaneous stationary reflection and some families of weak square principles. Finally we investigate the influence of the Rado's Conjecture on some polarized partition relations.

math.LO↗

arXiv2019

Computer-aided Detection of Squamous Carcinoma of the Cervix in Whole Slide Images

Goal: Squamous cell carcinoma of cervix is one of the most prevalent cancer worldwide in females. Traditionally, the most indispensable diagnosis of cervix squamous carcinoma is histopathological assessment which is achieved under microscope by pathologist. However, human evaluation of pathology slide is highly depending on the experience of pathologist, thus big inter- and intra-observer variability exists. Digital pathology, in combination with deep learning provides an opportunity to improve the objectivity and efficiency of histopathologic slide analysis. Methods: In this study, we obtained 800 haematoxylin and eosin stained slides from 300 patients suffered from cervix squamous carcinoma. Based on information from morphological heterogeneity in the tumor and its adjacent area, we established deep learning models using popular convolution neural network architectures (inception-v3, InceptionResnet-v2 and Resnet50). Then random forest was introduced to feature extractions and slide-based classification. Results: The overall performance of our proposed models on slide-based tumor discrimination were outstanding with an AUC scores > 0.94. While, location identifications of lesions in whole slide images were mediocre (FROC scores > 0.52) duo to the extreme complexity of tumor tissues. Conclusion: For the first time, our analysis workflow highlighted a quantitative visual-based slide analysis of cervix squamous carcinoma. Significance: This study demonstrates a pathway to assist pathologist and accelerate the diagnosis of patients by utilizing new computational approaches.

cs.CV↗

arXiv2019

TurboNet: A Model-driven DNN Decoder Based on Max-Log-MAP Algorithm for Turbo Code

This paper presents TurboNet, a novel model-driven deep learning (DL) architecture for turbo decoding that combines DL with the traditional max-log-maximum a posteriori (MAP) algorithm. To design TurboNet, we unfold the original iterative structure for turbo decoding and replace each iteration by a deep neural network (DNN) decoding unit. In particular, the DNN decoding unit is obtained by parameterizing the max-log-MAP algorithm rather than replace the whole decoder with a black box fully connected DNN architecture. With the proposed architecture, the parameters can be efficiently learned from training data, and thus TurboNet learns to appropriately use systematic and parity information to offer higher error correction capabilities and decrease computational complexity compared with existing methods. Furthermore, simulation results prove TurboNet's superiority in signal-to-noise ratio generalizations.

eess.SP↗

arXiv2019

Recent Advances in Transfer Learning for Cross-Dataset Visual Recognition: A Problem-Oriented Perspective

This paper takes a problem-oriented perspective and presents a comprehensive review of transfer learning methods, both shallow and deep, for cross-dataset visual recognition. Specifically, it categorises the cross-dataset recognition into seventeen problems based on a set of carefully chosen data and label attributes. Such a problem-oriented taxonomy has allowed us to examine how different transfer learning approaches tackle each problem and how well each problem has been researched to date. The comprehensive problem-oriented review of the advances in transfer learning with respect to the problem has not only revealed the challenges in transfer learning for visual recognition, but also the problems (e.g. eight of the seventeen problems) that have been scarcely studied. This survey not only presents an up-to-date technical review for researchers, but also a systematic approach and a reference for a machine learning practitioner to categorise a real problem and to look up for a possible solution accordingly.

cs.CV↗

arXiv2019

Experimental generation of phase wraps for subwavelength phase structures in Bose-Einstein condensate with two-dimensional optical lattice

We report an experimental demonstration of engineering phase wraps for sub-wavelength structure in a Bose-Einstein condensate (BEC) with two-dimensional optical lattices. A short lattice pulse is applied on BEC working in the Kapitza-Dirac (or Raman-Nath) regime, which corresponds to phase modulation imprint on matter wave. When the phase modulation on matter wave is larger than $2π$ in a lattice cell, there appears phase wraps with multiple $2π$ jumps, generating the sub-wavelength phase structure. The phase wraps for sub-wavelength structure are measured in momentum space via the time-of-flight absorption image, which corresponds to converting phase information into amplitude. %Two different kinds of two-dimensional optical lattice are studied, which show the different subwavelength phase structure. Moreover, we identify an additional condition for the validity of Kapitza-Dirac regime, which relies crucially on the lattice configurations. This scheme can be used for studying the property of optical lattices and topological defects in matter wave.

cond-mat.quant-gas↗

arXiv2019

Artificial Intelligence-aided Receiver for A CP-Free OFDM System: Design, Simulation, and Experimental Test

Orthogonal frequency division multiplexing (OFDM), usually with sufficient cyclic prefix (CP), has been widely applied in various communication systems. The CP in OFDM consumes additional resource and reduces spectrum and energy efficiency. However, channel estimation and signal detection are very challenging for CP-free OFDM systems. In this paper, we propose a novel artificial intelligence (AI)-aided receiver (AI receiver) for a CP-free OFDM system. The AI receiver includes a channel estimation neural network (CE-NET) and a signal detection neural network based on orthogonal approximate message passing (OAMP), called OAMP-NET. The CE-NET is initialized by the least-square channel estimation algorithm and refined by a linear minimum mean-squared error neural network. The OAMP-NET is established by unfolding the iterative OAMP algorithm and adding several trainable parameters to improve the detection performance. We first investigate their performance under different channel models through extensive simulation and then establish a real transmission system using a 5G rapid prototyping system for an over-the-air (OTA) test. Based on our study, the AI receiver can estimate time-varying channels with a single training phase. It also has great robustness to various imperfections and has better performance than those competitive algorithms, especially for high-order modulation. The OTA test further verifies its feasibility to real environments and indicates its potential for future communications systems.

cs.IT↗

arXiv2019

Deep Learning Based on Orthogonal Approximate Message Passing for CP-Free OFDM

Channel estimation and signal detection are very challenging for an orthogonal frequency division multiplexing (OFDM) system without cyclic prefix (CP). In this article, deep learning based on orthogonal approximate message passing (DL-OAMP) is used to address these problems. The DL-OAMP receiver includes a channel estimation neural network (CE-Net) and a signal detection neural network based on OAMP, called OAMP-Net. The CE-Net is initialized by the least square channel estimation algorithm and refined by minimum mean-squared error (MMSE) neural network. The OAMP-Net is established by unfolding the iterative OAMP algorithm and adding some trainable parameters to improve the detection performance. The DL-OAMP receiver is with low complexity and can estimate time-varying channels with only a single training. Simulation results demonstrate that the bit-error rate (BER) of the proposed scheme is lower than those of competitive algorithms for high-order modulation.

eess.SP↗

arXiv2019

Preparation of bipartite bound entangled Gaussian states in quantum optics

The positivity of the partial transpose is in general only a necessary condition for separability. There exist quantum states that are not separable, but nevertheless are positive under partial transpose. States of this type are known as bound entangled states meaning that these states are entangled but they do not allow distillation of pure entanglement by means of local operations and classical communication (LOCC). We present a parametrization of a class of $2\times 2$ bound entangled Gaussian states for bipartite continuous-variable quantum systems with two modes on each side. We propose an experimental protocol for preparing a particular bound entangled state in quantum optics. We then discuss the robustness properties of this protocol with respect to the occupation number of thermal inputs and the degrees of squeezing.

quant-ph↗

arXiv2019

Few-Shot Learning via Saliency-guided Hallucination of Samples

Learning new concepts from a few of samples is a standard challenge in computer vision. The main directions to improve the learning ability of few-shot training models include (i) a robust similarity learning and (ii) generating or hallucinating additional data from the limited existing samples. In this paper, we follow the latter direction and present a novel data hallucination model. Currently, most datapoint generators contain a specialized network (i.e., GAN) tasked with hallucinating new datapoints, thus requiring large numbers of annotated data for their training in the first place. In this paper, we propose a novel less-costly hallucination method for few-shot learning which utilizes saliency maps. To this end, we employ a saliency network to obtain the foregrounds and backgrounds of available image samples and feed the resulting maps into a two-stream network to hallucinate datapoints directly in the feature space from viable foreground-background combinations. To the best of our knowledge, we are the first to leverage saliency maps for such a task and we demonstrate their usefulness in hallucinating additional datapoints for few-shot learning. Our proposed network achieves the state of the art on publicly available datasets.

cs.CV↗

arXiv2019

Progressive LiDAR Adaptation for Road Detection

Despite rapid developments in visual image-based road detection, robustly identifying road areas in visual images remains challenging due to issues like illumination changes and blurry images. To this end, LiDAR sensor data can be incorporated to improve the visual image-based road detection, because LiDAR data is less susceptible to visual noises. However, the main difficulty in introducing LiDAR information into visual image-based road detection is that LiDAR data and its extracted features do not share the same space with the visual data and visual features. Such gaps in spaces may limit the benefits of LiDAR information for road detection. To overcome this issue, we introduce a novel Progressive LiDAR Adaptation-aided Road Detection (PLARD) approach to adapt LiDAR information into visual image-based road detection and improve detection performance. In PLARD, progressive LiDAR adaptation consists of two subsequent modules: 1) data space adaptation, which transforms the LiDAR data to the visual data space to align with the perspective view by applying altitude difference-based transformation; and 2) feature space adaptation, which adapts LiDAR features to visual features through a cascaded fusion structure. Comprehensive empirical studies on the well-known KITTI road detection benchmark demonstrate that PLARD takes advantage of both the visual and LiDAR information, achieving much more robust road detection even in challenging urban scenes. In particular, PLARD outperforms other state-of-the-art road detection models and is currently top of the publicly accessible benchmark leader-board.

cs.CV↗

arXiv2019

Towards Practical Visual Search Engine within Elasticsearch

In this paper, we describe our end-to-end content-based image retrieval system built upon Elasticsearch, a well-known and popular textual search engine. As far as we know, this is the first time such a system has been implemented in eCommerce, and our efforts have turned out to be highly worthwhile. We end up with a novel and exciting visual search solution that is extremely easy to be deployed, distributed, scaled and monitored in a cost-friendly manner. Moreover, our platform is intrinsically flexible in supporting multimodal searches, where visual and textual information can be jointly leveraged in retrieval. The core idea is to encode image feature vectors into a collection of string tokens in a way such that closer vectors will share more string tokens in common. By doing that, we can utilize Elasticsearch to efficiently retrieve similar images based on similarities within encoded sting tokens. As part of the development, we propose a novel vector to string encoding method, which is shown to substantially outperform the previous ones in terms of both precision and latency. First-hand experiences in implementing this Elasticsearch-based platform are extensively addressed, which should be valuable to practitioners also interested in building visual search engine on top of Elasticsearch.

cs.CV↗

arXiv2019

MirrorGAN: Learning Text-to-image Generation by Redescription

Generating an image from a given text description has two goals: visual realism and semantic consistency. Although significant progress has been made in generating high-quality and visually realistic images using generative adversarial networks, guaranteeing semantic consistency between the text description and visual content remains very challenging. In this paper, we address this problem by proposing a novel global-local attentive and semantic-preserving text-to-image-to-text framework called MirrorGAN. MirrorGAN exploits the idea of learning text-to-image generation by redescription and consists of three modules: a semantic text embedding module (STEM), a global-local collaborative attentive module for cascaded image generation (GLAM), and a semantic text regeneration and alignment module (STREAM). STEM generates word- and sentence-level embeddings. GLAM has a cascaded architecture for generating target images from coarse to fine scales, leveraging both local word attention and global sentence attention to progressively enhance the diversity and semantic consistency of the generated images. STREAM seeks to regenerate the text description from the generated image, which semantically aligns with the given text description. Thorough experiments on two public benchmark datasets demonstrate the superiority of MirrorGAN over other representative state-of-the-art methods.

cs.CL↗

arXiv2019

Attributes-aided Part Detection and Refinement for Person Re-identification

Person attributes are often exploited as mid-level human semantic information to help promote the performance of person re-identification task. In this paper, unlike most existing methods simply taking attribute learning as a classification problem, we perform it in a different way with the motivation that attributes are related to specific local regions, which refers to the perceptual ability of attributes. We utilize the process of attribute detection to generate corresponding attribute-part detectors, whose invariance to many influences like poses and camera views can be guaranteed. With detected local part regions, our model extracts local features to handle the body part misalignment problem, which is another major challenge for person re-identification. The local descriptors are further refined by fused attribute information to eliminate interferences caused by detection deviation. Extensive experiments on two popular benchmarks with attribute annotations demonstrate the effectiveness of our model and competitive performance compared with state-of-the-art algorithms.

cs.CV↗

arXiv2019

Almost global solutions to two classes of 1-d Hamiltonian Derivative Nonlinear Schrödinger equations

Consider two kinds of 1-d Hamiltonian Derivative Nonlinear Schrödinger (DNLS) equations with respect to different symplectic forms under periodic boundary conditions. The nonlinearities of these equations depend not only on $(x,ψ,\barψ)$ but also on $(ψ_x,\barψ_x)$, which means the nonlinearities of these equations are unbounded. Suppose that the nonlinearities depend on the space-variable $x$ periodically. Under some assumptions, for most potentials of these two kinds of Hamiltonian DNLS equations, if the initial value is smaller than $\varepsilon\ll1$ in $p$-Sobolev norm, then the corresponding solution to these equations is also smaller than $2\varepsilon$ during a time interval $(-c\varepsilon^{-r_*},c\varepsilon^{-r_*})$(for any given positive $r_*$). The main methods are constructing Birkhoff normal forms to two kinds of Hamiltonian systems which have unbounded nonlinearities and using the special symmetry of the unbounded nonlinearities of Hamiltonian functions to obtain a long time estimate of the solution in $p$-Sobolev norm.

math.DS↗