Steganalysis of neural networks based on parameter statistical bias

Many pretrained deep learning models have been released to help engineers and researchers develop deep learning-based systems or conduct research with minimall effort. Previous work has shown that at secret message can be embedded in neural network parameters without compromising the accuracy of the model. Malicious developers can, therefore, hide malware or other baneful information in pretrained models, causing harm to society. Hence, reliable detection of these vicious pretrained models is urgently needed. We analyze existing approaches for hiding messages and find that they will ineluctably cause biases in the parameter statistics. Therefore, we propose steganalysis methods for steganography on neural network parameters that extract statistics from benign and malicious models and build classifiers based on the extracted statistics. To the best of our knowledge, this is the first study on neural network steganalysis. The experimental results reveal that our proposed algorithm can effectively detect a model with an embedded message. Notably, our detection methods are still valid in cases where the payload of the stego model is low.     

基于无人机实时数据多波次任务规划模型分析

针对导弹部队多波次作战任务规划问题, 依据无人机的实时数据, 构建了基于路径的多层规划模型, 并设计了模型的算法求解流程。使用遗传算法与禁忌搜索混合算法, 得出了任务规划中的最优路径规划, 并在此基础上进行了冲突的消除。通过仿真案例表明, 用无人机协同配合导弹部队作战, 实时传输作战数据, 能够解决战场信息模糊不确定的问题; 使用多层规划模型能够为导弹多波次规划作战的路径进行科学的决策和选择。利用遗传算法和禁忌搜索混合算法, 能够避免局部最优导致无法输出结果的现象。     

Effect of Co substitution on the structural,dielectric and optical properties of KBiFe2O5

Cobalt (Co)-modified brownmillerite KBiFe2O5 (KBFO; [KBiFe2(1?x)Co2xO5 (x = 0, 0.05)]) polycrystalline is synthesized following the solid-state reaction route. Rietveld refinement of X-ray diffraction data confirmed the phase purity of KBFO and KBiFe1.9Co0.1O5 (KBFCO). The optical bandgap energy (Eg) of KBFO decreased from 1.59 to 1.51 eV because of Co substitution. The decrease in bandgap can be attributed to the tilting of the Fe–O tetrahedral structure of KBFCO. The observed room-temperature Raman peaks of KBFCO shifted by 3 cm?1 toward a lower wavenumber than that of KBFO. The shift in Raman active modes can be attributed to the change in the bond angles and bond lengths of the Fe–O tetrahedral structure and modification in response to oxygen deficiency in KBFO because of Co doping. Compared with that of KBFO, the frequency-dependent dielectric constant and dielectric loss of KBFCO decrease at room temperature, which is a con-sequence of the reduction in oxygen migration and modification in response to vibrational modes present in the sample.     

The Effectiveness of Structural Monetary Policy and Macro-Prudential Policies — Based on the DSGE Model That Includes Bank Heterogeneous Credit

Journal of Systems Science and Complexity - Structural monetary policy and macro-prudential policies are important parts of the policy system of the People’s Bank of China. By constructing a...     

多维度深海鱼类识别算法

针对深海光线分布不均匀导致鱼类识别检测困难的问题,提出了符合视觉认知的多维度深海鱼类识别算法.该方法从时间维度优化传统的高斯混合模型(GMM)初步确定变化区域,从空间维度构建目标特征,完整提取运动目标,从时空关联维度建立深度学习的鱼类识别框架,试验结果表明:本算法可在多种复杂条件下准确提取运动目标,面积交迭度(AOM)达到80％以上,优于当前主流算法.     

化学实验教学创新设计的调查报告

为了了解化学实验教学创新设计的现状,对常德市普通高中化学教师进行问卷调查,调查结果显示高中化学实验教学创新设计有弱化的趋势.在此基础上,从学科教学的角度对弱化的原因进行了分析,并提出化学实验教学创新设计的OCPP评价模式,尽可能建构、揭示或阐述实验教学创新设计的基本原理.OCPP模式包括目标评价(Objective Evaluation)、内容评价(Content Evaluation)、过程评价(Process Evaluation)和成果评价(Product Evaluation)4个部分.     

Dielectric relaxation and conduction mechanism in Aurivillius ceramic Bi5Ti3FeO15

For this study, we synthesized Aurivillius Bi5Ti3FeO15 ceramic using the generic solid-state reaction route and then performed room-temperature X-ray diffraction to confirm that the compound had a single phase with no impurities. The surface morphology of the prepared sample was observed to contain microstructural grains approximately 0.2–2 μm in size. The dielectric properties of the sample were determ-ined as a function of frequency in a range of approximately 100 Hz to 1 MHz at various temperatures (303 K ≤ T ≤ 773 K). Nyquist plots of the impedance data were found to exhibit a semi-circular arc in the high-temperature region, which is explained by the equivalent electrical circuit (R1C1)(R2QC2), where R1 and R2 represent the resistances associated with the grains and grain boundaries, respectively, C1 and C2 are the re-spective capacitances, and Q is the constant phase element (CPE), which accounts for non-Debye type of behavior. Our results indicate that both the resistance and capacitance of the grain boundaries are more prominent than those of the grains. The alternating current (ac) conductiv-ity data were analyzed based on the Jonscher universal power law, which indicated that the conduction process is dominated by the hopping mechanism. The calculated activation energies of the relaxation and conduction processes were very similar (0.32 to 0.53 eV), from which we conclude that the same type of charge carriers are involved in both processes.     

科技成果转化的探索与思考 ——基于湖南省科技成果转化的实证研究

科技成果转化是推动科技与经济结合的直接动力,是促进经济社会高质量发展的必要途径.文章通过对科技成果转化的重要性、现状和障碍的分析,提出科技行政管理部门应采取的管理举措.