Julie Dítětová: Machine Learning: Programming Patterns

In her presentation, artist and graphic designer Julie Dítětová explores the development of a neural network that generates ecclesiastical patterns. The lecture coincides with the opening of the exhibition of the same name.

This project utilizes the StyleGAN 2 neural network to generate visual data inspired by motifs from 18th-century church textiles. Initially, a digital archive of these textiles was established. Subsequently, the neural network was employed to create new patterns inspired by these historical motifs. The ensuing process involves iterative steps where the artist's programmed artificial intelligence experiments with different combinations of visual elements to accurately mimic the original patterns.