Mesh Fitter | 2-Dimensional Self-Organizing Maps using Crow in Grasshopper

New Grasshopper Plugin - Crow0.2 - Demo
This videos shows the use of the plug-in Crow for Grasshopper in Rhino5. It employs artificial neural networks - specifically a 2-dimensional Kohonen layer, known as self-organizing maps - to fit data vectors of arbitrary dimensions through a set of given training data. Since data in this example is comprised of position data of a point cloud and the network topology can be directly translated into a mesh structure, it is used to fit (tubular) meshes through point clouds.
Crow is an extension of NeuronDotNet.dll developed by the NeuronDotNetTeam

Get Crow here:
http://www.felbrich.com/projects/Crow/Crow.html