题目一:波达方向估计的低秩嵌入理论与方法
内容简介:目标测向又称波达方向估计,是雷达侦测的基础。本报告针对现有波达方向估计算法在困难场景下精度不足的问题,提出了一系列基于低秩嵌入的数学模型和算法。该方法从极大似然估计出发,通过将信号嵌入到结构性低秩矩阵中,精确刻画了雷达信号中常见的单通道、多通道和恒模等几何结构,将具有严重非凸性的原参数域优化问题等价刻画为信号域中的结构性低秩矩阵恢复问题,为原非凸优化问题求解提供了全新思路,带来了波达方向估计算法精度的本质提升。最后,本报告探讨了未来可能的研究方向。
报告人:杨在
报告人简介:西安交通大学数学与统计学院教授、博士生导师。2007和2009年分获中山大学应用数学本科和硕士学位,2014年获新加坡南洋理工大学博士学位。主要从事信息处理与无线通信的数学理论与方法研究,解决了Carathéodory-Fejér定理高维形式、两奇异半正定矩阵Hadamard积的正定性判定等公开问题,在IEEE Trans. Inf. Theory、IEEE Trans. Signal Process.、Appl. Comput. Harmonic Anal.等期刊与会议发表学术论文60余篇,谷歌学术引用3600余次。任或曾任IEEE Trans. Signal Process.期刊编委、欧洲信号处理会议Tutorial主讲人、IEEE信号处理学会传感器阵列与多通道(SAM)技术委员会委员。主持4项国家基金委项目(青年、两个面上及优青项目)、1项科技部重点研发课题,及多项华为企业横向课题。
题目二:A Direct Method for Calculating the Differential Spectrum of an APN Power Mapping
内容简介:Let $n$ be a positive integer, $p$ be an odd prime, $d=\frac{p^n+1}{4}+\frac{p^n-1}{2}$ if $p^n \equiv3 ~({\rm mod ~8})$ and $d=\frac{p^n+1}{4}$ if $p^n \equiv7 ~({\rm mod ~8})$. When $p^n>7$, the power mapping $x^d$ from $\mathbb{F}_{p^n}$ to $\mathbb{F}_{p^n}$ was proved to be almost perfect nonlinear by Helleseth, Rong and Sandberg in IEEE Trans. Inform. Theory, 45(2): 475-485, 1999. By establishing a system of linear equations related to the differential spectrum, Tan and Yan completely determined the differential spectrum of this power mapping in Des. Codes Cryptogr., 91(8): 2755-2768, 2023. In this talk, we will introduce another method of determining the differential spectrum of this APN function. We directly characterize the conditions on $b\in \mathbb{F}_{p^n}$ for which the differential $(x+1)^d-x^d=b$ has exactly $i$ solution(s) for $i=0,1,2$, respectively. Then, using the theory of elliptic curves, the number of those $b$'s in each case is determined and thus the differential spectrum of $x^d$ is obtained. Our method releases more information about the differential equation of $x^d$, which can be used to describe the DDT of this APN power function.
报告人:夏永波
报告人简介:中南民族大学教授,主持国家自然科学基金面上项目2项,国家自然科学基金青年项目1项,湖北省自然科学基金面上项目2项;在《IEEE Transactions on Information Theory》、《Finite Fields and Their Applications》、《Cryptography and Communications》、《中国科学数学(英文版)》等期刊上发表论文30余篇。曾获2018年湖北省自然科学奖二等奖(排名2)、2018年湖北省教学成果奖三等奖(排名3)、2019年国家民委教学成果奖二等奖(排名1),2019年入选国家民委青年教学标兵,2020年入选国家民委中青年英才培养计划。
题目三:Proximal linearized alternating direction method of multipliers algorithm for nonconvex image restoration with impulse noise
内容简介:Image restoration with impulse noise is an important task in image processing. Taking into account the statistical distribution of impulse noise, the $\ell_1$-norm data fidelity and total variation ($\ell_1TV$) model has been widely used in this area. However, the $\ell_1TV$ model usually performs worse when the noise level is high. To overcome this drawback, several nonconvex models have been proposed. In this paper, we propose an efficient iterative algorithm to solve nonconvex models arising in impulse noise. Compared to existing algorithms, our proposed algorithm is a completely explicit algorithm in which every subproblem has a closed-form solution. The key idea is to transform the original nonconvex models into an equivalent constrained minimization problem with two separable objective functions, where one is differentiable but nonconvex. As a consequence, we employ the proximal linearized alternating direction method of multipliers to solve it. We present extensive numerical experiments to demonstrate the efficiency and effectiveness of the proposed algorithm.
报告人:唐玉超
报告人简介:广州大学数学与信息科学学院,教授。主要研究方向图像处理中的优化模型和算法及其应用。在研国家自然科学基金项目和省杰出青年科学基金项目各1项,主持完成国家自然科学基金地区项目和国家自然科学基金青年项目各1项。已在《CSIAM Transactions on Applied Mathematics》、《Journal of Scientific Computing》,《Inverse Problems and Imaging》、《Set-Valued and Variational Analysis》和《中国科学数学》等国内外知名期刊发表SCI收录论文30余篇。中国数学会和中国工业与应用数学学会会员。美国数学评论员(112437)。2016年9月—2017年9月,受国家留学基金委资助在美国北卡罗来纳大学教堂山分校访问研究一年。
时 间:2024年3月22日(周五)上午9:30开始
地 点:南海楼124室
热烈欢迎广大师生参加!
信息科学技术学院/网络空间安全学院
2024年3月21日