Deformable Mesh
Date Issued
2005
Date
2005
Author(s)
Wu, Tai-Hei
DOI
en-US
Abstract
Neural Mesh is used to model deformable objects in this work. A modified Hopfield neural network is proposed to corporate with Neural Mesh. The characteristic of the modified Hopfield neural network is used to minimize the energy function of Neural Mesh. The energy function of Neural Mesh is approximated by a second-order Taylor series expansion. Several experiment results are given in the end of this work.
Reviews to related works, conclusions and discussions are given, too.
Subjects
變形網
類神經網路
Neural Mesh
Hopfield model
Neural Network
Deformable Objects
Type
thesis