ID 原文 译文
39056 但是现存的室内定位通常需要在离线阶段大量实验以建立相应特征库，有些系统还对设备有特殊要求。 However, existed device-free localization systems either suffer from offline training to establish corresponding feature library or require dedicated special-purpose devices.
39057 另外，现有的定位系统多是基于对二维平面上的运动进行定位，未曾考虑对进行三维运动的人进行定位。 Otherwise, most of the existing positioning systems was based on the positioning of the motion on the two-dimensional plane, no consideration had been given to tracking people who was walking in three-dimensional.
39058 对此本文提出了一种基于商用WiFi路由器和无线网卡下的系统，对人体反射回的无线信号做处理，从而得到多普勒频移、信号飞行时间、到达角度这些多维参数，对进行三维运动的人进行跟踪定位。 To this end, we proposed a system based on commercial WiFi router and wireless network card to track people who was walking in three-dimensional. Firstly, we processed the wireless signal reflected back by human body to figure out the multiple signal parameters including Angle-of-Arrival, Time-of-Flight, and Doppler shifts together. Secondly we devised an algorithm to tracking the pedestrian 3 d trajectory used the multiple parameters.
39059 本系统只用两个接收端，可以动态的适应环境，提取出多维用以估计位置的参数，达到0.7 m精度的三维轨迹跟踪。 Our system only used two link, can adapt to the environment dynamically, extracted the multi-dimensional parameters to tracking three-dimensional trajectory, and was able to achieve a accuracy below 0.7 m when the human is walking in three-dimensional.
39060 遥感图像目标检测能为军事和民用领域提供重要的可利用信息，成为近年来的研究热点。 Remote sensing image target detection can provide important usable information for military and civil fields and has become a research hotspot in recent years.
39061 针对现有目标检测技术不能兼顾检测速度和精度的问题，本文对Faster R-CNN做了优化:将轻量化的深度可分离残差网络作为Faster R-CNN的基础网络，降低基础网络模型的参数数量; Aiming at the problem that the existing target detection technology cannot give attention to both detection speed and accuracy, this paper optimized Faster R-CNN: the lightweight depth separable residual network was used as the backnone network of Faster R-CNN to reduce the number of parameters in the backnone network model.
39062 将基础网络中的多层卷积特征经局部响应归一化后进行融合，增强目标特征信息的完备性，改善小目标易漏检的问题; The multi-layer convolution features in the backnone network were fused after local response normalization to enhance the completeness of target feature information and improve the problem that small targets are easy to miss detection.
39063 联合softmax损失函数和中心损失函数训练网络模型，增加类别之间的差异性，缩小类内变化，使网络模型能学习到更具差异性的目标特征。 The network model was trained by combining softmax loss function and center loss function to reduce the changes within categories and increase the differences between categories, so that the network model can learn more different target features.
39064 在VEDAI、NWPU VHR-10、DOTA三个数据集上对本文方法进行验证，与传统Faster R-CNN相比，本文方法的检测精度提高了约7.0%。 The proposed method was verified on VEDAI, NWPU VHR-10, DOTA datasets. Compared with the traditional Faster R-CNN, the detection accuracy of the proposed method is improved by 7.0%.
39065 航空集群作战场景中电磁环境差且需传输业务量大，数据链频谱资源愈发紧张。 The poor electromagnetic environment and the large amount of traffic to be transmitted in the aviation cluster combat scenario have led to an increasingly tight spectrum of data link resources.