Search

Automatic classification of sea ice and open water plays a vital role in climate change research, polar shipping, and other applications. Many deep-learning-based methods are proposed to automatically classify sea ice and open water to address this issue. Even though these methods have achieved remarkable success, the noise phenomenon in synthetic aperture radar (SAR) images still causes considerable limitations in the model performance. Meanwhile, these existing methods ignore multiscale global information from large-scale SAR images, which tends to produce misclassification. In this article, we propose a novel multiscale dual attention network (MSDA-Net) for the task. To tackle the first drawback, we introduce the information of relative position and high-pass filtering as two extra channels to reduce the noisy effects. Moreover, we propose a patch dual attention mechanism and embed it into the ConvNeXt blocks to capture the multichannel and spatial features. To address the second problem, we propose a multiscale spatial attention module to capture multiscale global spatial information. The experiments show that the proposed method significantly outperforms state-of-the-art methods. In addition, comprehensive case studies are conducted, which verify the effectiveness of MSDA-Net in different SAR scenes.

Severe convection is a disastrous mesoscale weather system. The early detection of such systems is very important for saving peoples’ lives and properties. Previous studies address the issue mainly based on thresholding methods, which are not robust and accurate enough. In this paper, we propose a novel semantic segmentation method (Dugs-UNet) to solve the problem. Our method is based on the well-known U-Net framework. As convective clouds mimic fluids, its detection faces two important challenges. First, the shape and boundary features of clouds need to be carefully exploited. Second, the positive and negative samples for convection detection are very imbalanced. To address the two challenges, our method was carefully developed. Regarding the importance of the shape and boundary features for convective target detection, we introduce a shape stream module to extract these features. Also, a data-dependent upsample operation is adopted in the decoder of U-Net to effectively utilize the features. This is one of our contributions. To address the imbalance issue for convective target detection, the a focal loss function is employed to train our method, which is another contribution. Experimental results of 2018 Fengyun-4A satellite observations in China demonstrate the effectiveness of the proposed method. Compared to conventional thresholding-based methods and deep semantic segmentation algorithms such as SegNet, PSPNet, DeepLav-v3+ and U-Net, the proposed approach performs the best.

In recent years, silica nanomaterials have been widely studied as carriers in the field of antibacterial activity in food. Therefore, it is a promising but challenging proposition to construct responsive antibacterial materials with food safety and controllable release capabilities using silica nanomaterials. In this paper, a pH-responsive self-gated antibacterial material is reported, which uses mesoporous silica nanomaterials as a carrier and achieves self-gating of the antibacterial agent through pH-sensitive imine bonds. This is the first study in the field of food antibacterial materials to achieve self-gating through the chemical bond of the antibacterial material itself. The prepared antibacterial material can effectively sense changes in pH values caused by the growth of foodborne pathogens and choose whether to release antibacterial substances and at what rate. The development of this antibacterial material does not introduce other components, ensuring food safety. In addition, carrying mesoporous silica nanomaterials can also effectively enhance the inhibitory ability of the active substance.

Precipitation nowcasting is an important task, which can serve numerous applications such as urban alert and transportation. Previous studies leverage convolutional recurrent neural networks (RNNs) to address the problem. However, they all suffer from two inherent drawbacks of the convolutional RNN, namely, the lack of a memory cell to preserve the fine-grained spatial appearances and the position misalignment issue when combining current observations with previous hidden states. In this article, we aim to overcome the defects. Specifically, we propose a novel pseudo flow spatiotemporal LSTM unit (PFST-LSTM), where a spatial memory cell and a position alignment module are developed and embedded in the structure of LSTM. Upon the PFST-LSTM units, we develop a new sequence-to-sequence architecture for precipitation nowcasting, which can effectively combine the spatial appearances and motion information. Extensive empirical evaluations are conducted on synthetic MovingMNIST++ and CIKM AnalytiCup 2017 datasets. Our experimental results demonstrate the superiority of the proposed PFST-LSTM over the state-of-the-art competitors. To reproduce the results, we release the source code at: https://github.com/luochuyao/PFST-LSTM.

The Covid-19 pandemic is still spreading around the world and seriously imperils humankind's health. This swift spread has caused the public to panic and look to scientists for answers. Fortunately, these scientists already have a wealth of data—the Covid-19 reports that each country releases, reports with valuable spatial-temporal properties. These data point toward some key actions that humans can take in their fight against Covid-19. Technically, the Covid-19 records can be described as sequences, which represent spatial-temporal linkages among the data elements with graph structure. Therefore, we propose a novel framework, the Interaction-Temporal Graph Convolution Network (IT-GCN), to analyze pandemic data. Specifically, IT-GCN introduces ARIMA into GCN to model the data which originate on nodes in a graph, indicating the severity of the pandemic in different cities. Instead of regular spatial topology, we construct the graph nodes with the vectors via ARIMA parameterization to find out the interaction topology underlying in the pandemic data. Experimental results show that IT-GCN is able to capture the comprehensive interaction-temporal topology and achieve well-performed short-term prediction of the Covid-19 daily infected cases in the United States. Our framework outperforms state-of-art baselines in terms of MAE, RMSE and MAPE. We believe that IT-GCN is a valid and reasonable method to forecast the Covid-19 daily infected cases and other related time-series. Moreover, the prediction can assist in improving containment policies.

Abstract The unmanned aerial vehicle (UAV) can extend the network coverage and improve the system throughput for 5th generation (5G) communication systems; hence, it receives a lot of attention recently. This paper considers the problem of channel predictive precoding for UAV-enabled cache-assisted B5G multi-input multi-output (MIMO) systems. A novel channel precoder predictor is proposed, in which the prediction is conducted on a non-linear vector space—Grassmannian manifold. The predictor at the receiver utilizes the current and previous channel matrices to solve the precoder at the next time and then feeds it back to the transmitter for precoding. More specifically, two sub-matrices are extracted from the channel right singular matrices and modeled as two points on the Grassmannian manifold. Then, the geodesic between the two points is conducted. Unlike the conventional method in which the tangent vector at the previous point is parallel transported along the geodesic, we predict the next point by use of the geodesic equation directly. We analyze the computational complexity of the proposed method and demonstrate the superiority of the proposed method by comparing with the conventional one. Besides, we adopt a general Ricean channel model in the UAV MIMO system, where both the Kronecker model and Jake’s model are incorporated. The effects of various channel model parameters on the system performance in terms of the chordal error of channel predictor and the optimum step are thoroughly investigated.

This paper investigates the multi-relay non-orthogonal multiple access (NOMA) networks with cache-enabled users, in which the source transmits superimposed signals to two users, assisted by N relays. We firstly analyze traditional cooperative-NOMA networks, and then discuss the cooperative-NOMA networks with cache-enabled users. Moreover, the analytical expression of outage probability for the proposed network is derived, as well as the asymptotic expression. According to the asymptotic outage probability, we obtain that the diversity order is equal to the amount of relays N. Furthermore, increasing the file hit probability can enhance the transmission performance of the system. Finally, numerical as well as simulation results are presented to validate our analysis.

In the MIMO broadcasting system, channel state information (CSI) is often used for data detection at the receiver or preprocessing techniques such as the power control and user scheduling at the transmitter and hence, the study of its acquisition is important. Usually, the CSI can be obtained by the training-based channel estimation. Assuming that the users have the channels' statistical information and adopt the minimum mean-square-error (MMSE) receivers, this paper studies how to design an optimal training sequence to minimize the total weighted mean square error (MSE) of channel estimation. Following this objective, we formulate this problem as a positive semidefinite problem (SDP) with two constraints, and further transform it into a dual problem. In the solving process of the dual problem, we model it as an optimization problem on the positive definite matrix manifold and solve it by use of the iterative geodesic equation on the matrix manifold. The convergence and correctness of the proposed method are demonstrated by computer simulation; the effects of key system parameters on the weighted MSE are also investigated. Under various system configurations, the superiority of the proposed method is shown by comparing with a few other existing schemes.

Concurrent use of multiple drugs can lead to unexpected adverse drug reactions. The interaction between drugs can be confirmed by routine in vitro and clinical trials. However, it is difficult to test the drug–drug interactions widely and effectively before the drugs enter the market. Therefore, the prediction of drug–drug interactions has become one of the research priorities in the biomedical field. In recent years, researchers have been using deep learning to predict drug–drug interactions by exploiting drug structural features and graph theory, and have achieved a series of achievements. A drug–drug interaction prediction model SmileGNN is proposed in this paper, which can be characterized by aggregating the structural features of drugs constructed by SMILES data and the topological features of drugs in knowledge graphs obtained by graph neural networks. The experimental results show that the model proposed in this paper combines a variety of data sources and has a better prediction performance compared with existing prediction models of drug–drug interactions. Five out of the top ten predicted new drug–drug interactions are verified from the latest database, which proves the credibility of SmileGNN.

In this paper, we study a decode-and-forward (DF) relaying network with random interferers, in which the source transmits its message to the destination with the help of N DF relays. We consider the interference-limited environments, where the reception at the relays and the destination are corrupted by random interferers, which are distributed according to a homogeneous Poisson point process. To improve the system performance, relay selection based on received signal-to-interference ratio (SIR) has been employed to select the best relay among N ones. We examine the network performance by deriving the analytical outage probability under Rayleigh fading transmission channels and Nakagami-m fading interference channels. Moreover, we compute the asymptotic expressions of outage probability, and confirm that the system diversity order on SIR is equal to 2/α, where α denotes the path loss factor. Furthermore, we see that the major limitation of system results from the second hop. Numerical and simulation results are demonstrated to validate the proposed analysis as well.

The task of precipitation nowcasting is significant in the operational weather forecast. The radar echo map extrapolation plays a vital role in this task. Recently, deep learning techniques such as Convolutional Recurrent Neural Network (ConvRNN) models have been designed to solve the task. These models, albeit performing much better than conventional optical flow based approaches, suffer from a common problem of underestimating the high echo value parts. The drawback is fatal to precipitation nowcasting, as the parts often lead to heavy rains that may cause natural disasters. In this paper, we propose a novel interaction dual attention long short-term memory (IDA-LSTM) model to address the drawback. In the method, an interaction framework is developed for the ConvRNN unit to fully exploit the short-term context information by constructing a serial of coupled convolutions on the input and hidden states. Moreover, a dual attention mechanism on channels and positions is developed to recall the forgotten information in the long term. Comprehensive experiments have been conducted on CIKM AnalytiCup 2017 data sets, and the results show the effectiveness of the IDA-LSTM in addressing the underestimation drawback. The extrapolation performance of IDA-LSTM is superior to that of the state-of-the-art methods.

Cable termination is a weak point in an underground cable system. The transient earth voltage (TEV) method is an effective and nonintrusive method for estimating the insulation condition of cable termination. However, the practical application of TEV detection is mainly focused on switchgears, generators, and transformers with a flat and conductive shell. A flexible sensor array based on the TEV method is presented for online partial discharge (OLPD) monitoring of the cable termination. Each sensing element is designed with a dual-capacitor structure made of flexible polymer material to obtain better and more stable sensitivity. Based on the electromagnetic (EM) wave propagation theory, the partial discharge (PD) propagation model in the cable termination is built to analyze and verify the rationality and validity of the sensor unit. Some influencing factors are discussed regarding the response characteristics of sensors. Finally, the performance of the sensor array is verified by simulations and experiments. Besides, an OLPD monitoring system is introduced. The monitoring system is composed of the on-site monitoring device and the remote monitoring host. The two parts of the system exchange the data through wireless networks using a wireless communication module. The experiment results show that the monitoring device could supply the PD condition monitoring demand for cable termination.

Accurate hyperspectral image classification has been an important yet challenging task for years. With the recent success of deep learning in various tasks, 2-dimensional (2D)/3-dimensional (3D) convolutional neural networks (CNNs) have been exploited to capture spectral or spatial information in hyperspectral images. On the other hand, few approaches make use of both spectral and spatial information simultaneously, which is critical to accurate hyperspectral image classification. This paper presents a novel Synergistic Convolutional Neural Network (SyCNN) for accurate hyperspectral image classification. The SyCNN consists of a hybrid module that combines 2D and 3D CNNs in feature learning and a data interaction module that fuses spectral and spatial hyperspectral information. Additionally, it introduces a 3D attention mechanism before the fully-connected layer which helps filter out interfering features and information effectively. Extensive experiments over three public benchmarking datasets show that our proposed SyCNNs clearly outperform state-of-the-art techniques that use 2D/3D CNNs.

Presently, many users are involved in multiple social networks. Identifying the same user in different networks, also known as anchor link prediction, becomes an important problem, which can serve numerous applications, e.g., cross-network recommendation, user profiling, etc. Previous studies mainly use hand-crafted structure features, which, if not carefully designed, may fail to reflect the intrinsic structure regularities. Moreover, most of the methods neglect the attribute information of social networks. In this paper, we propose a novel semi-supervised network-embedding model to address the problem. In the model, each node of the multiple networks is represented by a vector for anchor link prediction, which is learnt with awareness of observed anchor links as semi-supervised information, and topology structure and attributes as input. Experimental results on the real-world data sets demonstrate the superiority of the proposed model compared to state-of-the-art techniques.