ID 原文 译文
1323 采用能量兼容设计理念，有助于应对电磁兼容的诱发因素，为规划、预防、处理设计中的电磁兼容问题提供必要条件。 Using the concept of energy compatibility design is helpful to deal with the inducing factors of EMC and provide requirements for planning, prevention and treatment of EMC problems in design.
1324 针对 SSD 原始附加特征提取网络(Original Additional Feature Extraction Network，OAFEN)中 stride 操作造成图像小目标信息丢失和串联结构产生的多尺度特征之间冗余度较大的问题，提出了一种计算量小、感受野大的深度可分离空洞卷积(Depthwise Separable Dilated Convolution，DSDC)， Aiming at the problems of small target information loss caused by stride operation and large redundancy a-mong multi-scale feature maps generated by serial structure in original additional feature extraction network (OAFEN)ofSSD, a depthwise separable dilated convolution (DSDC)with small computation and large field of receptivity is proposed;
1325 并利用 DSDC 设计了一个包含三个独立子网络的并行附加特征提取网络(Parallel Additional Feature Extraction Network，PAFEN)。 then a parallel additional feature extraction network (PAFEN)with three independent subnetworks is designed by using fiveDSDCs.
1326 PAFEN 上路用两个 DSDC 提取尺寸为19* 19 和 3* 3 的特征图; In upper subnetwork of PAFEN, two DSDCs are used to extract 19* 19 and 3* 3 feature maps.
1327 中路用一个 DSDC 提取尺寸为 10* 10 的特征图; In intermediate sub-network of PAFEN, one DSDC is used to extract 10* 10 feature maps.
1328 下路用两个 DSDC 提取尺寸为 5* 5 和 1*1 的特征图。 In lower subnetwork of PAFEN, two DSDCs are usedto extract 5* 5 and 1* 1 feature maps.
1329 实验结果表明，在 SSD 框架内，PAFEN 在 mAP 和检测时间等方面均优于 OAFEN，适用于地面小目标的检测任务。 The experimental results show that within the framework of SSD, PAFEN is superiorto OAFEN in terms of mAP and detection time, and is suitable for ground small target detection tasks.
1330 遥感影像检测分割技术通常需提取影像特征并通过深度学习算法挖掘影像的深层特征来实现。 Remote sensing image detection and segmentation technology usually needs to extract image features andmine the deep features of images through deep learning algorithm.
1331 然而传统特征(如颜色特征、纹理特征、空间关系特征等)不能充分描述影像语义信息，而单一结构或串联算法无法充分挖掘影像的深层特征和上下文语义信息。 However, traditional imaging features (e. g. , color, tex-ture, spatial relationship)cannot fully reflect the semantic information of the images, while single/sequential algorithm can-not fully exploit the deep features and the contextual semantic information of the images.
1332 针对上述问题，本文通过词嵌入将空间关系特征映射成实数密集向量，与颜色、纹理特征的结合。 Aiming at the above challenges, in this paper, the spatial relation features are mapped into real dense vectors by word embedding, which are combined with col-or and texture features.