Shape Space Estimation by SOM~2

摘要

This study aims to develop an estimation method for a shape space. In this work, 'shape space' is a nonlinear subspace formed by a class of visual shapes, in which the continuous change in shapes is naturally represented. By estimating the shape space, various operations dealing with shapes, such as identification, classification, recognition, and interpolation can be carried out in the shape space. A higher-rank of self-organizing map (SOM2) is employed as an implementation of the shape space estimation method. Simulation results show the capabilities of the method.