Simple jupyter notebook for generating abstract art with CPPN. Though techonolgy used in this notebook is CPPN, instead of using (x, y, latent vector) models gets (a, r, and optionally latent vector), where a stands for number of the rectangle pixel belongs from the edge of the image (for example all of the pixels will have 0, all of the pixels on the next rectangle from the edge will have 1, and etc.) and r for radius from the center of the image. This approach allows to create new images which do not completely look like basic CPPN. As well the model can generate animations with walking weights space (!not latent space!). It's done for a reason that updating latent vector with constant add-on will eventiually make the model to generate numbers exceeding 255, which will be round up to 255, and eventually image will look like white canvas. Using approch as walking sin function with random direction and step for every single weight will heli preventing the situation described above, as well as generating completely random images from the beggining till the end.
-
Notifications
You must be signed in to change notification settings - Fork 0
Generating abstract art with CPPN
License
ivpravdin/cppn-abstract-art
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Generating abstract art with CPPN
Topics
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published