ID 原文 译文
38946 为了提高标注的准确率，设置一个频率阈值筛选出频率较高的标签作为最终的关键词。 Technically, set a threshold of the word frequency to select the labels with the higher frequency as the final keywords.
38947 通过实验证明，实验结果有效地提高了标注的准确率和召回率。 More remarkably, the experimental results show that the method effectively improves precision and recall rate of the annotation.
38948 光响应非均匀性(photo-response non-uniformity，PRNU)是用于数字图像设备溯源的一种重要特征，也被称为成像设备指纹。 Photo-response non-uniformity(PRNU) referred as device fingerprint can be used for digital imaging device source identification.
38949 针对图像真实噪声包含PRNU和大量未知噪声的复杂特性，本文提出一种结合深度迭代缩放卷积神经网络的PRNU数字成像设备指纹提取算法。 In view of the complexity of PRNU and a large number of unknown noises in the real image noise, this paper proposed a PRNU extraction algorithm combined with deep iterative down-up convolutional neural network(CNN) model.
38950 首先，通过连续重复的缩小与放大特征图的分辨率来提高GPU内存利用效率和生成大的感受野，尽可能的提取包含完整PRNU指纹的真实噪声。 In particular, in order to extract the complete PSNR fingerprint from the real image noise, the resolution of the feature map was decreased and increased repetitively to efficiently employ graphics processing unit(GPU) memory and yield large receptive fields.
38951 然后，利用来自同一数字成像设备多幅图像的噪声残差来估计PRNU指纹。 Then, the noise residual was computed by multiple images from the same digital imaging device, which was utilized to estimate the PRNU fingerprint.
38952 本文算法在相机溯源数据集Dresden和手机溯源数据集Daxing上进行了测试。 The proposed algorithm was evaluated by Dresden camera database and Daxing smartphone database,
38953 实验结果显示，与基于小波变化的算法、基于块匹配3D滤波的算法以及基于前馈去噪卷积神经网络的算法相比，本文算法具有的更好的识别率和普适性。 and the experimental results show that the proposed algorithm achieves better recognition accuracy and universality compared with traditional methods.
38954 与基于可见光的单模态跟踪算法相比，基于可见光/热红外(RGB-Thermal， RGBT)的双模态跟踪算法对光照变化具有更强的鲁棒性。 Compared with single-modal RGB trackers, dual-modal RGBT(RGB-Thermal) trackers are more robust to illumination variation.
38955 但在实际应用场景中，双模态跟踪算法仍然会受到局部遮挡目标形变的影响。 In real scenarios, however, dual-modal trackers are severely influenced by partial occlusion and shape deformation.