ID 原文 译文
40836 RISC-V作为近年来发展迅速的开源处理器架构，对其验证方法的研究具有很高的工业应用价值。 As an open source processor architecture with rapid development in recent years, RISC-V has high industrial application value in the research of its verification method.
40837 提出了基于遗传进化策略的处理器验证优化方案，利用遗传算法产生指令的权重系数，为测试用例的生成确定方向，并对RISC-V处理器RV32I部分指令完成了模拟验证。 In this paper, an optimization scheme of processor verification based on the genetic evolution strategy is proposed, which uses a genetic algorithm to generate instruction weight coefficients to determine the direction for the generation of test cases, and some instructions of RV32 I of the RISC-V processor are simulated and verified.
40838 与基于伪随机策略生成测试用例的方案相比，应用遗传进化策略生成测试用例时，能够维持对功能覆盖率贡献更高的方向， Compared with the scheme based on the pseudo-random strategy to generate test cases, the application of genetic evolution strategy to generate test cases can maintain a higher contribution to the function coverage.
40839 在验证末期即将陷入局部最优解时，仍然能够突破当前状态，使整体功能覆盖率进一步提高。 When it is about to fall into a locally optimal solution at the end of verification, it is still able to break out of the current state, allowing a further improvement of the overall functional coverage.
40840 同时，该验证优化方案对验证环境的调整实现了完全自动化，具有较高的可扩展性和可维护性。 At the same time, the verification optimization scheme realizes the automatic adjustment of the verification environment and has high scalability and maintainability.
40841 为了防止攻击者在深度学习图像分类过程中还原训练集数据并保护输入图像数据，提出一种基于层级相关性传播的差分隐私分类算法。 To prevent attackers from restoring the training dataset and to protect the input image during the application process of deep learning image classification model, a differential privacy classification algorithm based on LRP(Layer-wise Relevance Propagation) is proposed in the paper.
40842 该算法首先采用层级相关性传播模型量化图像的特征相关性，然后利用相关性自适应地向损失函数添加噪声并利用Adam算法进行模型优化， The relevance between image features is firstly quantified according to LRP in the proposed algorithm, then adaptive noise is added to the loss function based on feature relevance and Adam mechanism is used for model optimization.
40843 最后依据相关性分配隐私预算并构造差分隐私变换层以扰动输入数据。 Finally, a differential private transform layer is constructed to perturb the input image and privacy budget is assigned according to feature relevance during model application stage.
40844 实验结果表明，该算法在实现隐私保护的同时，能够保证较高的分类准确率。 Experimental results show that the proposed algorithm achieves high classification accuracy in condition of privacy preservation.
40845 硬件安全已经成为当今时代备受瞩目的焦点问题，硬件木马检测的技术也逐渐变得多样化。 Hardware security has become the focus of attention in today′s era, hardware Trojans detection techniques are also becoming more diverse.