| ID | 原文 | 译文 |
| 53747 | 但是现有的对抗训练过程生成对抗样本时需要类别标记信息，并且会大大降低无攻击数据集上的泛化性能。 | Yet it requires semantic annotation information to generate adversarial attacks and often perform poorly on the original data set. |
| 53748 | 本文提出一种基于自监督对比学习的深度神经网络对抗鲁棒性提升方法， | This paper proposes a self-supervised contrastive learning framework to adversarially train a robust neural network without labeled data. |
| 53749 | 充分利用大量存在的无标记数据改善模型在对抗场景中的预测稳定性和泛化性。 | We aim to maximize the representation similarity between a random augmentation of an image and its instance level adversarial perturbation. |
| 53750 | 本文所提方法可以用于预训练模型的鲁棒性提升，也可以与对抗训练相结合最大化模型的“预训练+微调”鲁棒性，在遥感图像场景分类数据集上的实验结果证明了所提方法的有效性和灵活性。 | The proposed method can be used to improve the adversarial robustness of pre-trained models, and can also be used to enhance the two stage robustness. We validate the proposed method on two remote sensing scene classification benchmark data sets. |
| 53751 | 水下传感节点静默定位算法是一种免时钟同步且易拓展节点的水声定位算法，待定位节点只需接收声信号即可计算出自身位置。 | The silent location algorithm for underwater sensor nodes is a time-synchronization free location algorithm where nodes could join expediently. Nodes to be located can calculate themselves' positions just with monitoring sound signal. |
| 53752 | 受其独特的定位协议影响，其定位周期一般较长，定位精度受待定位节点自身运动影响较大。 | Due to the particular location protocol, the location period of the silent location algorithm is generally longer. So that the location accuracy is greatly affected by the motion of the nodes to be located. |
| 53753 | 本文研究了待定位节点运动对已有算法精度的影响，并提出了一种结合移动补偿的动态水下待定位节点静默定位方法，补偿了待定位节点运动对自身定位精度的影响。 | In this paper, how the motion effect the accuracy of existing silent location algorithm is investigated, and a silent location algorithm for dynamic underwater sensor nodes combined with mobility compensation is proposed to eliminate the motion of the node to be located. |
| 53754 | 待定位节点首先估计来自各信标节点信号到达时刻的变化状态，掌握其变化规律，并据此以同一接收位置为基准补偿得到的观测信息。 | In order to learn the change regular of measurements, the nodes to be located estimated the change state of the arrival time of signals from each beacon node firstly. Furthermore, the arrival time measured at the same positon could be predicted according to the estimated change regular. |
| 53755 | 随后结合已有静默定位算法获得自身位置。 | The motion-compensated positon could be calculated using the existing silent location algorithm with the predicted measurements. |
| 53756 | 仿真结果表明，本文所提算法可以有效地抑制待定位节点运动产生的定位误差。 | The effectiveness of the proposed algorithm in this paper is verified by simulation analysis. |