Aesthetic Evolution of Target 3D Volumes in Minecraft

Evolutionary Art is a type of generative art which evolves until it achieves the desired form. Our approach to Evolutionary Art is a generator in Minecraft that aims to recreate a cuboid from a hand-drawn approximation of five of its faces. We attempt to approximate the target $32 \times 32 \times 32$ cuboid from its provided face projections, and use edge detection to further extract possible “intent“ from the shape boundaries in the target images. Our aim was not to fully approximate the target; we could achieve this goal with relatively simple deterministic algorithms. We relied on the Genetic Algorithm’s global search as a source of “creativity”, and sought to obtain varied structures somewhat similar to our pre-drawn target image. Our approach managed to create structures resembling the target (or at least the parts of the target that are not highly conflicting) while struggling with one of the targets containing small details. The fitness function was fine-tuned, during many iterations, to strike the balance between faithful reproduction and stylistically-consistent exploratory departures from the target, according to the authors’ tastes. However, these results are purely subjective since art is, by itself, subjective. Our Genetic Algorithm works with variable-length chromosomes, containing a mixture of individual voxels and a list of larger geometric shapes. The operators are adapted both to the chromosomal representation and the 3-dimensional geometry they work in. Furthermore, the algorithm is integrated in a toolchain which is capable of taking a hand-drawn image from a graphical tablet and providing tens of fully-interactive volumes in Minecraft, with a waiting time of between 5 and 20 minutes.