Parameter Estimation of An Adapted Mean-curvature Flow Model of Shape Evolution

来源 :第八届工业与应用数学国际大会 | 被引量 : 0次 | 上传用户:NewComeMan
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  This work has been initiated through the visual analogy between the early cortical folding process and the smoothing of a brain surface by mean curvature flow.We introduce a new geometric flow with one-parameter and propose an efficient optimization strategy for parameter estimation involving an energy depending only on volume and total area of the closed surface.Our model is trained on brain data and reveals promising predictions for developmental neuroscience.
其他文献
  The coupled-cluster method is a highly accurate wavefunction method solving a many-body Schrodinger's equation.However,its computational complexity scales a
会议
  For strongly interacting electronic systems,the Kohn-Sham formulation of density-functional theory leads to an optimal transport problem with Coulomb cost f
会议
  Olfactory processing in the brain exhibits a number of rhythms,which signify enhanced coherence among many neurons.While the faster gamma rhythm is generate
会议
  The Electronic Structure Library(ESL)is a new initiative to create an online repository of software for use within electronic structure codes.One of the aim
会议
  This talk presents the development of a real space approach to perform efficient large-scale Kohn-Sham density functional theory calculations using an adapt
会议
  We use the information relaxation technique to develop a primaldual iterative approach to solve stochastic dynamic programming problems.In each iteration,we
会议
  The minisymposium concerns control of partial differential equations,analysis of stochastic systems and a population dynamic model.It is one of the series o
会议
  We consider networks of nonlinear strings and Timoshenko beams as well as Cosserat networks.We provide the modeling,the analysis of equilibrium solutions an
会议
  Backward SPDEs arise in many applications of probability theory and stochastic processes.The lack of the solution's regularity brings huge difficulty to stu
会议
  We present Lagrangean and Eulerian evolving manifold models and computational approaches used for 3D and 4D image segmentation,3D point cloud surface recons
会议