An approach based on neural networks for animation

Abstract

neural network-based a new method is In this study, presented for animation. For this purpose, 2D and 3D Initial boundary curves are given as meshes are used. inputs to single layer feed forward neural network and inner points are calculated as output values of neural Inner points are used for meshing process. network. Animation is implemented by changing inner points. In 2D, triangulation process is done by meshing inner points. To compute inner points, weight values of neural networks are used. Different inner points are obtained by changing weight values. Remeshing are made by changing weights values of neural networks on the generated mesh. And so, it is examined changes with respect to animation. Some figures are given as application in 2D.