Study of a Class of Almost‐controllable Discrete‐time Bilinear Systems

This paper studies the controllability of a class of time‐invariant discrete‐time bilinear systems. Although the system is not controllable in the whole space, there is a very large region where control is effective. Results show that the uncontrollable region of this kind of bilinear system has a Lebesgue measure of zero. In other words, for almost any initial state and any terminal state in the state space, the former can be transferred to the latter. Further, a necessary condition for near controllability is presented. Therefore, the results in this paper unify and generalize the corresponding conclusions in the literature. ©2014 Chinese Automatic Control Society and Wiley Publishing Asia Pty Ltd     

New (t, n) threshold directed signature scheme with provable security

Directed signature scheme allows only a designated verifier to check the validity of the signature issued to him; and at the time of trouble or if necessary, any third party can verify the signature with the help of the signer or the designated verifier as well. Due to its merits, directed signature scheme is widely used in situations where the receiver’s privacy should be protected. Threshold directed signature is an extension of the standard directed signature, in which several signers may be required to cooperatively sign messages for sharing the responsibility and authority. To the best of our knowledge, threshold directed signature has not been well studied till now. Therefore, in this paper, we would like to formalize the threshold directed signature and its security model, then present a new (t, n) threshold directed signature scheme from bilinear pairings and use the techniques from provable security to analyze its security.     

Globally stabilizing feedback control of process systems in generalized Lotka–Volterra form

In the present paper a globally stabilizing feedback controller design method is proposed for process systems when the feedback structure is also assumed to be in a quasi-polynomial form. It is shown that such feedback structure can always be achieved for process systems.By exploiting the special structure of the controller design problem, the existent iterative linear matrix inequality (ILMI) algorithm of [Y.-Y. Cao, J. Lam, Y.-X. Sun, Static output feedback stabilization: an ILMI approach, Automatica 12 (1998) 1641–1645] is applied to solve the BMI feasibility problem underlying the design.In addition, some partial results on placing the globally stable equilibrium point with respect to the positive orthant have also been proposed that is only possible in a fully actuated situation when the input variables are the intensive variables at the inlet. Furthermore, some preliminary results in selecting the structure of the QP-type feedback have also been presented.     

多线程管理下的数据库访问技术

对数据库服务器访问的管理,是提高C／S架构系统性能的重要环节。采用多线程技术对即时访问通道进行动态管理,可有效降低系统资源消耗,提高系统响应速度。     

基于视角置信度和注意力的暴力行为识别

暴力行为容易出现遮挡情况, 识别准确率较低. 目前, 一些算法加入多视角视频输入来解决遮挡问题, 以等量权重将所有视角数据融合, 但是不同视角的视频因拍摄距离和遮挡情况本身就对识别存在差异性. 针对该问题, 本文提出一种基于视角置信度和注意力的暴力行为识别方法, 提高暴力识别的准确率. 本文将时序差分模块TDM的输入扩展成多视角, 将通道注意力机制运用在片段维度来增强TDM中跨段特征提取能力, 通过背景抑制方法突显移动目标的纹理特征并计算出每个视角图像的置信度, 引入双线性池化方法融合多视角视频特征, 根据视角置信度分配每个视角局部特征的权重. 本文在公开数据集CASIA-Action和自制数据集上进行了验证. 实验表明, 本文提出的视角置信度方法优于改进前的双线性池化方法, 暴力行为准确率相较于现有的行为识别方法取得了更好的效果.     

FSNet: 基于频率特性的烟雾图像分割网络

烟雾图像分割是对烟雾进行识别与精准定位的基础, 是火灾预警的重要手段. 针对烟雾分割时存在过分割、欠分割以及边界拟合粗糙的问题, 本文提出一种基于频率分离特性的烟雾图像分割网络. 所提出的频率分离模块将特征图中的烟雾区域分离为低频主体部分和高频边界部分, 同时基于多任务学习设计多模块权重自适应损失函数对烟雾整体、主体、边界分别监督学习, 起到细化烟雾边界和改善烟雾整体分割结果的作用; 此外, 结合可变形卷积提出改进的空洞空间金字塔池化模块以解决其信息利用率低和特征关联性差的问题. 在对比实验中, FSNet的烟雾交并比为76.55%, 比基线网络提高了4.25%. 可视化分割结果可以看出, FSNet能有效缓解过分割、欠分割, 所得烟雾边界更平滑, 烟雾图像分割的整体性能获得较大提升.     

基于超像素级卷积神经网络的多聚焦图像融合算法

该文提出了基于超像素级卷积神经网络(sp-CNN)的多聚焦图像融合算法。该方法首先对源图像进行多尺度超像素分割,将获取的超像素输入sp-CNN,并对输出的初始分类映射图进行连通域操作得到初始决策图;然后根据多幅初始决策图的异同获得不确定区域,并利用空间频率对其再分类,得到阶段决策图;最后利用形态学对阶段决策图进行后处理,并根据所得的最终决策图融合图像。该文算法直接利用超像素分割块进行图像融合,其相较以往利用重叠块的融合算法可达到降低时间复杂度的目的,同时可获得较好的融合效果。     

CCNet: CNN model with channel attention and convolutional pooling mechanism for spatial image steganalysis

Image steganalysis based on convolutional neural networks(CNN) has attracted great attention. However, existing networks lack attention to regional features with complex texture, which makes the ability of discrimination learning miss in network. In this paper, we described a new CNN designed to focus on useful features and improve detection accuracy for spatial-domain steganalysis. The proposed model consists of three modules: noise extraction module, noise analysis module and classification module. A channel attention mechanism is used in the noise extraction module and analysis module, which is realized by embedding the SE(Squeeze-and-Excitation) module into the residual block. Then, we use convolutional pooling instead of average pooling to aggregate features. The experimental results show that detection accuracy of the proposed model is significantly better than those of the existing models such as SRNet, Zhu-Net and GBRAS-Net. Compared with these models, our model has better generalization ability, which is critical for practical application.     

A pairing-free identity-based authenticated key agreement protocol with minimal message exchanges

Based on the computational Diffie-Hellman problem, this paper proposes an identity-based authenticated key agreement protocol which removes bilinear pairings. Compared with previous protocols, the new protocol minimizes message exchange time with no extra cost. The protocol provides strong security guarantees including key compromise impersonation resilience, perfect forward secrecy, and master key forward secrecy. A security proof with the modular approach in the modified Bellare-Rogaway model is also provided.     

一种安全有效的基于身份的门限签密方案

签密是一种将加密和数字签名技术结合在一起的思想,它比采用先签名后加密的方法具有更高的效率.基于双线性对技术,提出了一种新的基于身份的门限签密方案.它利用决策双线性Diffie-Hellman(DBDH)问题的困难性在随机预言模型下给出了方案的语义安全性证明,并利用计算Diffie-Hellman(CDH)问题的困难性证明了方案的不可伪造性,同时指出方案具有较高的效率.