Learning Genetic Representations as Alternative to Hand-Coded Shape Grammars

Schnier, Thorsten; Gero, John S.

doi:10.1007/978-94-009-0279-4_3

Cited by 28 publications

(14 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shape grammars are especially useful in the context of design and architecture, since design heuristics can be coded into the grammar. Examples of the use of shape grammars in evolutionary art and design are [37] and [30]. [31] describes the use of shape grammars (using the Context Free language) to evolve multiple artworks in a similar style.…”

Section: Representation In Evolutionary Artmentioning

confidence: 99%

Using scalable vector graphics to evolve art

Heijer

Eiben

2016

IJART

View full text Add to dashboard Cite

In this paper we describe our investigations of the use of Scalable Vector Graphics as a genotype representation in evolutionary art. We describe the technical aspects of using SVG in evolutionary art, and explain the genetic operators mutation and crossover. Furthermore, we compare the use of SVG with existing representations in evolutionary art. We perform two series of experiments and describe their setup and results. In the first series of experiments we investigate the feasibility of SVG as a genotype representation for evolutionary art, and we evolve abstract images using a number of aesthetic measures as fitness functions (all experiments described in this paper are done without a human in the loop). We found that SVG is suitable as a genotype representation for evolutionary art, but that the range of the visual output was limited by the design of our genetic operators. In order to increase the range of the visual output, and in order to evolve representational images, we performed a second series of experiments in which we used existing images as source material. We designed and implemented a new initialisation, crossover and mutation operator. We also designed and implemented an ad-hoc aesthetic measure for 'pop-art' and used this to evolve images that are visually similar to screen printing art and pop art. All images and SVG code examples in this paper are available at

show abstract

Section: Representation In Evolutionary Artmentioning

confidence: 99%

Using scalable vector graphics to evolve art

Heijer

Eiben

2016

IJART

View full text Add to dashboard Cite

show abstract

“…Other development models have taken a "generative" approach, such as string grammars to develop house plants [64], or L-systems to encode plant morphologies [57,34], 3D branched organisms [7], objects made of voxels [28,27], and surfaces in 3D spaces [25]. Shape grammars [76], a production system to generate geometric shapes, have been evolved for architectural designs [67]. Grammatical evolution, a method to evolve productions of a string grammar by encoding the rules of the production in a binary genome [51], has been applied to shape grammars to evolve simple shapes [52].…”

Section: A Graph-grammar-based Developmental Model For Behavior-findingmentioning

confidence: 99%

Behavior-Finding: Morphogenetic Designs Shaped by Function

Lobo

Fernández

Vico

2012

Morphogenetic Engineering

View full text Add to dashboard Cite

Evolution has shaped an incredible diversity of multicellular living organisms, whose complex forms are self-made through a robust developmental process. This fundamental combination of biological evolution and development has served as an inspiration for novel engineering design methodologies, with the goal to overcome the scalability problems suffered by classical top-down approaches. Top-down methodologies are based on the manual decomposition of the design into modular, independent subunits. In contrast, recent computational morphogenetic techniques have shown that they were able to automatically generate truly complex innovative designs. Algorithms based on evolutionary computation and artificial development have been proposed to automatically design both the structures, within certain constraints, and the controllers that optimize their function. However, the driving force of biological evolution does not resemble an enumeration of design requirements, but much rather relies on the interaction of organisms within the environment. Similarly, controllers do not evolve nor develop separately, but are woven into the organism's morphology. In this chapter, we discuss evolutionary morphogenetic algorithms inspired by these important aspects of biological evolution. The proposed methodologies could contribute to the automation of processes that design "organic" structures, whose morphologies and controllers are intended to solve a functional problem. The performance of the algorithms is tested on a class of optimization problems that we call behavior-finding. These challenges are not explicitly based on morphology or controller constraints, but only on the solving abilities and efficacy of the design. Our results show that morphogenetic algorithms are well suited to behavior-finding.

show abstract

“…In this second example the genetic engineering approach is used to learn style features from a set of example floor plans, in this case from floor plans of Frank Lloyd Wright houses [19] for a more detailed description of this application, see. An analysis of the style used shows that some of the style features are not simply products of outlines , hut are linked to the functions of the rooms.…”

Section: Example 2: Learning Style Featuresmentioning

confidence: 99%

Genetic Engineering and Design Problems

Gero

Казаков

Schnier

1997

Evolutionary Algorithms in Engineering Applications

Self Cite

View full text Add to dashboard Cite

Summary. In this chapter we describe an extension of genetic algorithms which utilizes a set of ideas borrowed from the practice of genetic engineering of natural organisms. We show how it can be constructed and implemented computationally and how it enhances the efliciency ofthe genetic algorithm's computational process . We apply it to the solution of a spatial layout planning problem as an exarnple of an NP-complete problem and to case-based layout design as an example of a dass of problem which cannot be solved directly using traditional genetic algorithms.

show abstract

Learning Genetic Representations as Alternative to Hand-Coded Shape Grammars

Cited by 28 publications

References 12 publications

Using scalable vector graphics to evolve art

Using scalable vector graphics to evolve art

Behavior-Finding: Morphogenetic Designs Shaped by Function

Genetic Engineering and Design Problems

Contact Info

Product

Resources

About