This page presents a method of digital art generation that combines the evolution of Compositional Pattern Producing Networks (CPPNs) with Procedural Content Generation via Wave Function Collapse (WFC). CPPNs are a type of generative encoding that has been used to create a variety of art, including images, sculptures, and animations. WFC can arrange various images in a grid by solving an adjacency constraint problem. The two tools can be combined to create intricate patterns, which can be mixed together to create art reminiscent of Zentangles, a meditative art form. The system can generate art through both automatic evolution, and interactive evolution.
This system was developed by undergraduate students Anna Krolikowski, Sarah Friday, and Alice Quintanilla as a term project for the class CSC54-424 - Artificial Intelligence at Southwestern University in Spring 2019. Anna and Sarah then developed the system further for the purpose of publication. Examples of Zentangles created by the system are featured below. To get full access to the data from automated and interactive evolution experiments discussed in the paper, go to this Google Drive link. The source code for evolving these Zentangles is available on GitHub.
Video Presentation
Video presentation prepared by Anna Krolikowski and Sarah Friday for EvoMUSART 2020 after the conference shifted to a virtual format due to the COVID-19 pandemic. This video was actually a backup in case of technical difficulties with a synchronous presentation, but was not needed. This video was also used for the Southwestern University Research and Creative Works Symposium.
Human Interactive Evolution
Results of humans interactively evolving art for Zentangles.
Half-Black Fitness
Automatic evolution: Attempts to make images have 50% black pixels, so that final Zentangles have a mesh of different patterns.
Half-Black-3-Colors Fitness
Automatic evolution: Like Half-Black, but also has separate fitness functions for each color channel (RGB) that are competing to maximize presence of that color. Creates diverse color palettes in population.
Random Fitness (Color)
Automatic evolution: Random fitness function with color images.
Random Fitness (Black and White)
Automatic evolution: Random fitness function with black-and-white images.
Jacob Schrum,
Pier Luca Lanzi,
Alexander Nareyek,
and Pieter Spronck
(2018).
Human-assisted Creation of Content Within Games,
Dagstuhl Reports: Artificial and Computational Intelligence in Games: AI-Driven Game Design (Dagstuhl Seminar 17471), Volume 7, No. 11, page 111. Editors: Pieter Spronck and Elisabeth André and Michael Cook and Mike Preuß. Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik
Fall
2018:
The cover of SIGEVOlution Volume 11, Issue 4 features art generated by AnimationBreeder, the interactive evolution system described in a GECCO 2018 paper co-authored with SU students. SIGEVO is the ACM Special Interest Group on Genetic and Evolutionary Computation.
Summer
2018:
Neuroevolution in Video Games: "Mad Science Monday" presentation
made by my SCOPE Summer research students to present to other SCOPE students