ID 原文 译文
40986 把白平衡图像和锐化图像作为输入，进行NSCT分解。 The white balance and sharpening images are taken as the input for NSCT decomposition.
40987 对于低频分量，提出了一种区域方差与加权拉普拉斯能量和的融合方法; For low frequency components, a fusion method of regional variance and weighted Laplace energy sum is proposed.An improved guided filter is used to process the high frequency components.
40988 采用了改进的引导滤波对高频分量进行处理; Finally, inverse NSCT transform and color correction are carried out to obtain clear underwater images.
40989 最后进行逆NSCT变换和颜色校正，获得清晰的水下图像。实验结果表明，该算法纠正了水下图像色偏，提升图像对比度，突出细节信息，显著提高了水下图像的质量。 Experimental results show that the algorithm corrects the color distortion, improves the image contrast, highlights the details, and significantly improves the quality of the underwater images.
40990 机器学习系统以其强大的自适应性、自学习能力，越来越多的应用到Android恶意软件检测领域，取得了显著的检测效果。 With its strong self-adaptability and self-learning ability, machine learning systems are increasingly used in the field of Android malware detection and have achieved remarkable detection results.
40991 然而，机器学习算法和样本本身还面临着诸多安全威胁，一些经过精心策划的攻击，希望颠覆这些算法并允许恶意行为对抗检测。 However, machine learning algorithms and samples themselves also face many security threats, some well-designed attacks, hoping to subvert these algorithms and allow malicious behavior to resist detection.
40992 首先以Drebin系统为例介绍了基于机器学习的Android恶意软件检测方法的原理，然后在攻击目标、攻击策略的基础了提出了针对机器学习分类器的逃避攻击模型。 Firstly, the principle of machine learning-based detection method is introduced by taking the Drebin system as an example, and then an evasion attack model for machine learning classifier is proposed based on the attack target and attack strategy.
40993 在综合考虑特征权重、可修改性、修改成本的基础上，提出了一种恶意对抗样本生成方法。 Based on the comprehensive consideration of feature weight, modifiability and modification cost, proposes a malicious countermeasure sample generation method.
40994 实验结果表明，只需要修改很少量的特征，就能够逃避线性SVM分类器的检测， The experimental results show that only a few features need to be modified to evade the detection of the linear SVM classifier.
40995 最后用一个具体的恶意样本逃避实例验证了提出方法的有效性。 Finally, a specific example of malicious sample evasion attack verifies the effectiveness of the proposed method.