本文提出一种基于双向生成对抗网络（Bidirectional Generative Adversarial Network， BiGAN）的无监督感知哈希生成算法，通过编码网络、生成网络和判别网络间的双向迭代对抗，生成具有较强图像语义特征表示能力的感知哈希码. 本算法通过在编码网络和生成网络间添加跳接层网络结构，将原始图像不同维度的特征信息传递到生成网络，提高生成图像语义学习能力与网络收敛速度；同时，在对抗损失中添加均方误差（Mean Sequare Error， MSE）损失，增强生成图像的视觉质量与细节表示能力. 最后，基于网络间的多重迭代对抗训练，输出兼备相同来源图像鲁棒性和不同来源图像区分性的高性能图像感知哈希码. 本研究首次采用大型图像数据库进行算法性能评价，实验结果表明基双向生成对抗网络的感知哈希生成算法与当前其他最新研究方案相比具有更强的版权认证与来源检测能力.

传统的感知哈希算法通过提取图像特定属性生成感知哈希序列,难以充分利用原始图像全部特征信息,影响了基于感知哈希的图像内容认证与版权保护能力.本文提出一种基于双向生成对抗网络(Bidirectional Genera-tive Adversarial Network,BiGAN)的无监督感知哈希图像内容取证算法,基于编码网络、生成网络和判别网络间的双向迭代对抗,生成具有较强图像语义特征表示能力的感知哈希码;并通过在编码网络和生成网络间添加跳接层网络结构,将原始图像不同维度的特征信息传递到生成网络,提高生成网络语义特征学习能力与网络收敛速度;同时,在对抗损失中添加MSE误差损失,增强生成图像的视觉质量与细节表示能力;最后,基于网络间的多重迭代与对抗训练,输出兼具相同内容图像认证鲁棒性和不同内容图像区分敏感性的高性能图像感知哈希码.本研究首次采用大型图像数据库进行算法性能评价,实验结果表明基于双向生成对抗网络的感知哈希图像内容取证算法与当前其他优秀研究方案相比具有更强的图像内容取证性能.

为强化立德树人目标,进一步推进课程思政高质量建设,促进专业课与思政理论课同向同行,实现知识目标、能力目标和素养目标的有机统一,结合网络空间安全专业和信息系统安全课程的特点,深入挖掘信息系统安全课程专业知识中蕴含的思政元素,重新设计知识、能力及育人三者同样重要的教学目标,从课程性质、教学目标、课程思政教学设计及内容、教学方法和教学效果等方面对课程进行设计和总结,旨在探索一条能够使专业学习与思政教育相互促进、协同发展的教改之路.

Most of the existing deep learning-based synchronous JPEG double compression detection algorithms use the truncation and rounding errors generated in the decompression process as the classification basis. Pre-processing layers that reduce the difficulty of feature extraction are present before the detection framework, and end-to-end detection cannot be achieved. Meanwhile, the existing quantization base table is designed based on human experience and cannot obtain the optimal solution for the decompression process, which limits the accuracy of the JPEG double compression detection algorithms. To address these issues, a JPEG double compression detection method based on a decompression module is proposed. The proposed method exploits convolution to simulate the JPEG decompression process, and designs the decompression module to incorporate the JPEG decompression process into the network to achieve end-to-end detection, which is free from laborious pre-processing steps. At the same time, the optimal solution for the decompression process is automatically searched based on the self-optimized characteristic of deep learning, which can reduce the secondary loss of image information caused by manual processing and further improve the performance of the JPEG double compression detection algorithm. The experimental results show that the proposed synchronous JPEG double compression detection algorithm achieves better forensic performance in more than half of the experimental groups, with an average accuracy improvement of up to 1.8% over the existing methods on the UCID dataset.

The rapid development of generative adversarial networks(GANs) has led to unprecedented success in the field of image generation.The emergence of new GANs such as StyleGAN makes the generated images more realistic and deceptive,posing a greater threat to national security,social stability,and personal privacy.In this paper,a detection algorithm based on a space-frequency joint two-stream convolutional neural network is proposed.Since GAN images will leave clearly discernible artifacts on the spectrum due to the up-sampling operation during the generation process,a learnable frequency-domain filter kernel and frequency domain network are designed to fully learn and extract frequency-domain features.In order to reduce the influence of the information discarded from the image transformation to the frequency domain,a spatial domain network is also designed to learn that the image content itself has differentiated spatial domain features.Finally,the two features are fused to detect the face image generated by GAN.Experimental results on multiple datasets show that the proposed model outperforms existing algorithms in detection accuracy on high-quality generated datasets and generalization across datasets.And for JPEG compression,random cropping,Gaussian blur,and other operations,this method has stronger robustness.In addition,the proposed method also performs well on the local face dataset generated by GAN,which further proves that this model has better generality and wider application prospects.

计算机网络是计算机类专业一门重要的核心课程。在线上教育迅速发展的背景下，探索应用雨课堂实施以学生学习需求为导向的贯通“课前线上预习—课中线上参与式学习—课后线上提高”的“三阶贯通”式线上教学模式，将所有教与学的活动通过一个平台串联，取得较好的教学效果。实践表明，此种线上教学模式可以为今后开展线上线下混合式教学提供范式。

人工智能的发展为信息隐藏技术带来越来越多的挑战,提高现有隐写方法的安全性迫在眉睫.为提高图像的信息隐藏能力,提出一种基于U-Net结构的生成式多重对抗隐写算法.所提算法通过生成对抗网络与隐写分析器优化网络、隐写分析对抗网络间的多重对抗训练,构建生成式多重对抗隐写网络模型,生成适合信息隐写的载体图像,提高隐写图像抗隐写分析能力;同时,针对现有生成对抗网络只能生成随机图像,且图像质量不高的问题,设计基于U-Net结构的生成式网络模型,将参考图像的细节信息传递到生成载体图像中,可控地生成高质量目标载体图像,增强信息隐藏能力;其次,采用图像判别损失、均方误差(MSE)损失和隐写分析损失动态加权组合作为网络迭代优化总损失,保障生成式多重对抗隐写网络快速稳定收敛.实验表明,基于U-Net结构的生成式多重对抗隐写算法生成的载体图像PSNR最高可达到 48.60 dB,隐写分析器对生成载体图像及其隐写图像的判别率为50.02%,所提算法能够生成适合信息嵌入的高质量载体图像,保障隐写网络快速稳定收敛,提高了图像隐写安全性,可以有效抵御当前优秀的隐写分析算法的检测.

目前国内众多网络空间安全学院都开设了"网络空间安全数学基础"课程,该课程在教学过程中还存在着一些共性问题:"重"理论知识传授,"轻"课堂引入;"重"计算和证明,"轻"动手实践能力培养;未能将课程思政元素融入课堂;采用"一考定成绩"的传统考核方式.分析这些问题的成因及危害,提出相应的对策:精心选择课堂引入方式;注重培养学生的动手实践能力;将课程思政有机地融入课堂教学;采用注重过程的考核方式.