Urban flooding is a major natural disaster that poses serious threat to the urban environment. It highly demanded flood extent can be mapped in near real-time for rescue and relief missions, reconstruction efforts, financial loss evaluation. Many efforts have been taken identify zones with remote sensing data image processing techniques. Unfortunately, production of accurate maps over impacted areas has not well investigated due three issues. (1) Satellite imagery high spatial resolution...

This paper proposes a deep Convolutional Neural Network(CNN) with strong generalization ability for structural topology optimization. The architecture of the neural network is made up encoding and decoding parts, which provide down- up-sampling operations. In addition, popular technique, namely U-Net, was adopted to improve performance proposed network. input well-designed tensor each channel includes different information problem, output layout optimal structure. To train network, large...

We present the data from 11 observing campaigns (carried out between 1989 and 1999) at Effelsberg 100 m radio telescope to study Intraday Variability in Active Galactic Nuclei. Most of these observations were performed total power linear polarization. give summary tables, light curves, structure functions sets. Due large number individual observations, only examples lightcurves will be presented here; complete set figures accessible online. variations are nearly all sources (detected during...

One of the key steps in pavement maintenance is fast and accurate identification distresses, defects, markings ability to conduct before irreversible damages. Recently, convolution neural network (CNN) has emerged as a powerful tool automatically identify cracks, where many CNN models take long computation time. To solve problem, an adaptive lightweight model, named MobileCrack, proposed this study. MobileCrack realizes using following settings: (1) reduce input image size, besides original...

A resource allocation scheme involving several Device-to-Device (D2D) multicast groups underlaying OFDMA-based systems is proposed in this paper. In order to guarantee Quality of service (QoS) cellular users (CeUEs), we define the signal-to-interference-pulse noise ratio (SINR) threshold value for CeUEs. Meanwhile, minimum throughput a single D2D group predetermined maintain fairness each group. Since subcarrier assignment and power are joint optimization problem with complex calculation,...

Near realtime flood mapping in densely populated urban areas is critical for emergency response. The strong heterogeneity of poses a big challenge accurate near mapping. However, previous studies on automatic methods perform infeasible or fail to generalize well other floods, several reasons. First, multitemporal pixel-wise requires image registration, hindering the efficiency large-scale processing. Although registration has been investigated, precisely coregistered sequence time-consuming...

Since the high randomness feature of traffic flows, it is challenging to be aware spatial-temporal characteristics dynamic vehicle loads on a long-span bridge. Compared with conventional measuring approaches, computer vision-based techniques have competitive advantages in measurement factual parameters, including passing time, velocity, type, and corresponding locations bridge at each timestamp. In this paper, framework for identifications bridges proposed. To specific, consists two major...

Reducing grid carbon emissions plays a crucial role in addressing climate change and achieving sustainable energy development. Accurate forecasting of intensity contributes to adjusting consumption strategies enhancing clean utilization. Due the uncertainty renewable output, often exhibits non-stationarity stochasticity. To this end, we propose probabilistic method based on time series decomposition. First, Hodrick–Prescott filter is studied extract different components from original...

Video summarization aims to extract keyframes/shots from a long video. Previous methods mainly take diversity and representativeness of generated summaries as prior knowledge in algorithm design. In this paper, we formulate video content-based recommender problem, which should distill the most useful content for users who suffer information overload. A scalable deep neural network is proposed on predicting if one segment by explicitly modelling both Moreover, accomplish scene action...

Advances in deep learning and computer vision are making significant contributions to flood mapping, particularly when integrated with remotely sensed data. Although existing supervised methods, especially convolutional neural networks, have proved be effective, they require intensive manual labeling of flooded pixels train a multi-layer network that learns abstract semantic features the input This research introduces novel weakly approach for pixel-wise mapping by leveraging multi-temporal...

We present the data of three observing campaigns with VLA performed to study intraday variability in compact extragalactic radio sources. The first campaign (May 1989) lasted 5 days; total intensity as well linear polarization were obtained at 2, 3.6, 6, and 20 cm wavelength for 9 program In second run (February 1990), a five-antenna subarray was used 25 days monitor sources 0716+714, 0917+624, 0954+658, dense sampling same four wavelengths. third (October 1992), which 22 days, 8 monitored...

This paper analyzes the optimal resource allocation for multi-D2D(Device-to-Device) links underlying OFDMA-based networks with target of reaching maximum sum-throughput D2D transmission. In order to guarantee QoS (Quality Service) cellular users, rate loss users is limited by a lower-bound constraint RLC. Meanwhile, link allowed use sub-carrier only when throughput from spectrum reusing exceeds certain threshold. Since primal problem non-convex, we solve it through subgradient-based...

Person identification in the wild is very challenging due to great variation poses, face quality, clothes, makeup and so on. Traditional research, such as recognition, person re-identification, speaker often focuses on a single modal of information, which inadequate handle all situations practice. Multi-modal more promising way that we can jointly utilize face, head, body, audio features, In this paper, introduce iQIYI-VID, largest video dataset for multi-modal identification. It composed...

Among the existing methods for maintenance and monitoring of bridges, human eye evaluation, which is inevitably subjective time-consuming, still most widely applied. In this paper, a new automatic inspection method deterioration bottom steel box girder based on computer vision proposed. First, system installed bridge vehicle used to capture photos girder, are synthesized into panoramas by image stitching technology. Then, U-net semantic segmentation network identify diseases in panoramas....

The Effelsberg 100 m radio telescope has been used, over a period of 5.5 years, to monitor the flux densities 40 extragalactic sources, including complete sample 13 flat spectrum sources from S5 Survey, and various other active sources. study their long-term variability characteristics at several wavelengths, source light curves, spectra, statistical analysis are presented discussed in this paper.

This paper investigates the time-domain oversampled cyclic-prefix orthogonal frequency division multiplexing (CP-OFDM) system for communication over doubly-selective underwater acoustic channels. Channel estimation is based on two different time-varying channel models. One basis expansion model (BEM), other Doppler-rate (DRM). Exploiting sparsity of channels, matching pursuit (OMP) employed to obtain state information (CSI). The receiver operates without any prior knowledge statistics CSI,...

Abstract A large parallel‐cable manipulator for the feed‐supporting system of a next‐generation radio telescope is presented in this paper. The approximate kinematics model developed to improve real‐time controllability, and rationality approximation validated by accuracy analysis. In order guarantee effectiveness control, singularity analyzed (including force singularities). control strategy also proposed. © 2001 John Wiley & Sons, Inc.

Deep learning has been applied in various fields for its effective and accurate feature capabilities recent years. Currently, information extracted from remote sensing images with the methods become most relevant research area developed precision. In terms of developing segmentation precision reducing calculation power consumption, improved deep have received more attention, improvement semantic architectures a popular solution. This presents method named D-DenseNet new structure road...

This paper analyzes the effect of cooperation on network capacity in hybrid composed cellular and D2D communications. The communication shares uplink resource with user we consider a simple system model which contains one pair user. First derive out cumulative distribution function (CDF) interference at receiver dual channel. Then study based former corollary different scenarios according to CSI acknowledgement transceiver case by case. simulation results show that between link can bring...

Real-time processing of anomaly detection has become one the most important issues in hyperspectral remote sensing. Due to fact that widely used imaging spectrometers work a pushbroom fashion, it is necessary process incoming data line causal linewise progressive manner with no future involved. In this study, we proposed several processes well improve computational performance real-time (RCLP-AD). At first, Cholesky decomposition along linear system solving (CDLSS) was since background...