ID	原文	译文
57068	该框架围绕交通工具运行安全问题的特性——"高速运动状态下网络资源的可用性、信息物理融合下安全目标的复杂性",通过精准感知物理层、通信层、社会层3个层面的不安全事件,综合利用泛在智能感知、边缘–云计算等技术,从交通工具的物理设备信息安全、车辆状态安全、车辆环境安全、网络安全4个维度,构建一体化的交通工具智能安全应对方案,服务于"事前全时预判、事中即时报警、事后回放取证"的安全目标.	The proposed framework aims to resolve the issues of mobile vehicle security, includingthe availability of network resources in high-speed motion and the complexity of security objectives within cyber?physical systems. Relying on precise perception of insecure events at the physical, communication, and societylayers, this paper constructs an integrated intelligent safety response strategy for physical equipment informationsecurity, state vehicle security, environmental vehicle security, and network security by intelligent perception,edge-cloud computing, and other technologies. The proposed framework achieves the goals of real-time eventprediction before the event, immediate alarm during the event, and replay for evidence forensics after the event.
57069	雷达探测是一种有效的对地观测手段,雷达图像目标识别是其重要研究方向.	Radar detection is an effective earth observation means, and target recognition in radar images isits important research direction.
57070	深度学习已在众多领域取得成功,然而训练深层神经网络需要大量数据,样本量不足已成为制约深度学习方法在雷达图像目标识别中应用的主要因素.	Deep learning has been successfully applied to many fields but training deepneural networks requires a mass of data. The lack of samples has become the major factor that impedes theapplication of deep learning approaches to target recognition in radar images.
57071	本文对基于深度学习的雷达图像目标识别研究进展进行了综述,梳理和总结了具有代表性的方法.	This paper reviews the researchprogress of deep learning based target recognition in radar images, with representative methods being combedand summarized.
57072	首先,介绍针对雷达图像目标识别的数据扩充和网络模型设计方法.	First, data augmentation and neural network models designed for the task of radar image targetrecognition are introduced.
57073	然后,详细阐述本课题组针对小样本条件提出的基于迁移学习、度量学习和半监督学习的雷达图像目标识别方法.	The paper then presents in detail the target recognition methods based on transferlearning, metric learning, and semi-supervised learning in radar images with few samples, which are proposed byour research group.
57074	最后,讨论了目前仍然存在的问题,并给出了对未来发展趋势的展望.	Finally, existing problems and future development trends are discussed.
57075	建筑物三维重建在城市规划、灾害监测、智慧城市等领域有重要应用,是计算机视觉、摄影测量、遥感等领域研究的重要课题.	Owing to its momentous applications in urban planning, disaster monitoring, smart cities and otherfields, 3D building reconstruction is an important topic in many research areas such as computer vision, pho?togrammetry, and remote sensing.
57076	由于SAR成像机理的特殊性和复杂性,基于SAR图像的建筑物三维重建难度很大,现有方法的适用性和自动化程度都亟待提升.	The particularity and complexity of the microwave scattering mechanism bringgreat challenges to the 3D building reconstruction of SAR images, and the applicability and automation of existingmethods need to be improved. This study constructs the overall framework of building detection in SAR imagesand 3D reconstruction based on deep learning and radar imaging mechanism.
57077	本文构建了基于深度学习与雷达成像机理结合的SAR图像建筑物检测及三维重建整体框架,并提出了基于耦合等效复数卷积神经网络的SAR图像建筑侧立面检测方法,基于RaySAR的建模仿真及点云生成方法,以及基于3D生成网络的SAR建筑物三维重建方法,利用TerraSAR-X与GF-3高分辨率SAR图像进行实验,得到了较好的三维重建结果.	It puts forward a method of usinga coupled equivalent complex valued convolutional neural network for building facade detection in SAR images, amethod for RaySAR-based modeling simulation and point cloud generation for 3D model training, and a methodfor a 3D generation network for 3D building reconstruction from SAR images. Experiments using TerraSAR-Xand GF-3 high resolution SAR images are carried out, producing good 3D reconstruction results.