ITU

A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations

Research output: Conference Article in Proceeding or Book/Report chapterArticle in proceedingsResearchpeer-review

Standard

A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations. / Risi, Sebastian.

2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research.. 2013.

Research output: Conference Article in Proceeding or Book/Report chapterArticle in proceedingsResearchpeer-review

Harvard

Risi, S 2013, A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations. in 2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research..

APA

Risi, S. (2013). A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations. In 2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research.

Vancouver

Risi S. A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations. In 2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research.. 2013

Author

Risi, Sebastian. / A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations. 2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research.. 2013.

Bibtex

@inproceedings{c11fcdfb635141d2abfcc33f45ea838d,
title = "A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations",
abstract = "Biologically-inspired AI methods like evolutionary algorithms have shown great promise in creating complex structures yet these structures still pale in comparison to their natural counterparts. The recently introduced generative encoding compositional pattern producing networks (CPPNs), which is based on the principles of how natural organisms develop, narrowed this gap by showing that it is possible to artificially evolve life-like patterns with regularities at a high-level of abstraction. As these generative and developmental systems (GDS) are asked to evolve increasingly complex structures, the question of how to start evolution from a promising part of the search space becomes more and more important. To address this challenge, we introduce the concept of a CPPN-Compiler, which allows the user to directly compile a high-level description of the desired starting structure into the CPPN itself. In this paper, as proof of concept, the CPPN-Compiler is able to generate CPPN-encoded representations from vector-based images that can serve as the starting point for further evolution. Importantly, the offspring of these compiled CPPNs show meaningful variations because they directly embody important domain-specific regularities like symmetry or repetition. Thus the results presented in this paper open up a new research direction in GDS, in which specialized CPPN-Compilers for different domains could help to overcome the black box of evolutionary optimization.",
author = "Sebastian Risi",
note = "ISBN mangler. Det er en symposieserie der udkommer hver halve {\aa}r. Muligvis udkommer serien kun elektronisk p{\aa} nettet. Jeg unders{\o}ger n{\ae}rmere. 4.1.2014 haal Stadig ikke publiceret. 22.1.2014 haal",
year = "2013",
language = "English",
isbn = "978-1-57735-640-0",
booktitle = "2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research.",

}

RIS

TY - GEN

T1 - A Compiler for CPPNs: Transforming Phenotypic Descriptions Into Genotypic Representations

AU - Risi, Sebastian

N1 - ISBN mangler. Det er en symposieserie der udkommer hver halve år. Muligvis udkommer serien kun elektronisk på nettet. Jeg undersøger nærmere. 4.1.2014 haal Stadig ikke publiceret. 22.1.2014 haal

PY - 2013

Y1 - 2013

N2 - Biologically-inspired AI methods like evolutionary algorithms have shown great promise in creating complex structures yet these structures still pale in comparison to their natural counterparts. The recently introduced generative encoding compositional pattern producing networks (CPPNs), which is based on the principles of how natural organisms develop, narrowed this gap by showing that it is possible to artificially evolve life-like patterns with regularities at a high-level of abstraction. As these generative and developmental systems (GDS) are asked to evolve increasingly complex structures, the question of how to start evolution from a promising part of the search space becomes more and more important. To address this challenge, we introduce the concept of a CPPN-Compiler, which allows the user to directly compile a high-level description of the desired starting structure into the CPPN itself. In this paper, as proof of concept, the CPPN-Compiler is able to generate CPPN-encoded representations from vector-based images that can serve as the starting point for further evolution. Importantly, the offspring of these compiled CPPNs show meaningful variations because they directly embody important domain-specific regularities like symmetry or repetition. Thus the results presented in this paper open up a new research direction in GDS, in which specialized CPPN-Compilers for different domains could help to overcome the black box of evolutionary optimization.

AB - Biologically-inspired AI methods like evolutionary algorithms have shown great promise in creating complex structures yet these structures still pale in comparison to their natural counterparts. The recently introduced generative encoding compositional pattern producing networks (CPPNs), which is based on the principles of how natural organisms develop, narrowed this gap by showing that it is possible to artificially evolve life-like patterns with regularities at a high-level of abstraction. As these generative and developmental systems (GDS) are asked to evolve increasingly complex structures, the question of how to start evolution from a promising part of the search space becomes more and more important. To address this challenge, we introduce the concept of a CPPN-Compiler, which allows the user to directly compile a high-level description of the desired starting structure into the CPPN itself. In this paper, as proof of concept, the CPPN-Compiler is able to generate CPPN-encoded representations from vector-based images that can serve as the starting point for further evolution. Importantly, the offspring of these compiled CPPNs show meaningful variations because they directly embody important domain-specific regularities like symmetry or repetition. Thus the results presented in this paper open up a new research direction in GDS, in which specialized CPPN-Compilers for different domains could help to overcome the black box of evolutionary optimization.

M3 - Article in proceedings

SN - 978-1-57735-640-0

BT - 2013 AAAI Fall Symposium Series - How Should Intelligence be Abstracted in AI Research.

ER -

ID: 62966260