ID 原文 译文
19515 最后采用残差学习的方法来提高网络的性能。 Finally, the proposed network uses residual learning to improve network performance.
19516 为了提高网络的通用性，采用具有不同压缩质量因子的联合训练方式对网络进行训练，针对不同压缩质量因子训练出一个通用模型。 In order to improve the versatility of the network, the network is trained by a joint training method with different compression quality factors, and a general model is trained for different compression quality factors.
19517 经实验表明，该文方法不仅具有较高的JPEG压缩伪迹去除性能，且具有较强的泛化能力。 Experiments demonstrate that the proposed algorithm not only has high JPEG compression artifacts reduction performance, but also has strong generalization ability.
19518 在遥感图像语义分割中，利用多元数据(如高程信息)进行辅助是一个研究重点。 Utilizing multiple data (elevation information) to assist remote sensing image segmentation is an important research topic in recent years.
19519 现有的基于多元数据的分割方法通常直接将多元数据作为模型的多特征输入，未能充分利用多元数据的多层次特征， However, the existing methods usually directly use multivariate dataas the input of the model, which fails to make full use of the multi-level features.
19520 此外，遥感图像中目标尺寸大小不一，对于一些中小型目标，如车辆、房屋等，难以做到精细化分割。 In addition, the target sizevaries in remote sensing images, for some small targets, such as vehicles, houses, etc., it is difficult to achievedetailed segmentation.
19521 针对以上问题，提出一种多特征图金字塔融合深度网络(MFPNet)，该模型利用光学遥感图像和高程数据作为输入，提取图像的多层次特征， Considering these problems, a Multi-Feature map Pyramid fusion deep Network(MFPNet) is proposed, which utilizes optical remote sensing images and elevation data as input to extractmulti-level features from images.
19522 然后针对不同层次的特征，分别引入金字塔池化结构，提取图像的多尺度特征， Then the pyramid pooling structure is introduced to extract the multi-scalefeatures from different levels.
19523 最后，设计了一种多层次、多尺度特征融合策略，综合利用多元数据的特征信息，实现遥感图像的精细化分割。 Finally, a multi-level and multi-scale feature fusion strategy is designed, which utilizes comprehensively the feature information of multivariate data to achieve detailed segmentation of remote sensing images.
19524 基于Vaihingen数据集设计了相应的对比实验，实验结果证明了所提方法的有效性。 Experiment results on the Vaihingen dataset demonstrate the effectiveness of the proposed method.