Modern Evolution Strategies for Creativity: Fitting Concrete Images and Abstract Concepts
This work addresses computational creativity for digital art and AI applications, offering incremental improvements by applying modern ES to an existing domain.
The paper tackled the problem of using evolutionary algorithms for computational creativity by applying modern evolution strategies (ES) to optimize shape placement, resulting in large improvements in quality and efficiency compared to traditional genetic algorithms and competitive performance with gradient-based methods, including producing diverse geometric abstractions aligned with human language interpretation.
Evolutionary algorithms have been used in the digital art scene since the 1970s. A popular application of genetic algorithms is to optimize the procedural placement of vector graphic primitives to resemble a given painting. In recent years, deep learning-based approaches have also been proposed to generate procedural drawings, which can be optimized using gradient descent. In this work, we revisit the use of evolutionary algorithms for computational creativity. We find that modern evolution strategies (ES) algorithms, when tasked with the placement of shapes, offer large improvements in both quality and efficiency compared to traditional genetic algorithms, and even comparable to gradient-based methods. We demonstrate that ES is also well suited at optimizing the placement of shapes to fit the CLIP model, and can produce diverse, distinct geometric abstractions that are aligned with human interpretation of language. Videos and demo: https://es-clip.github.io/