Shallow Art: Art Extension Through Simple Machine Learning

TL;DR

This paper explores using simple machine learning models for art extension, training on various image collections to generate completed artworks and examining their implications for computational creativity.

Contribution

It introduces a novel approach of using shallow classification and regression models for art generation and extension tasks.

Findings

Models can successfully complete missing parts of images.

Simple models produce plausible art extensions.

Implications for computational creativity are discussed.

Abstract

Shallow Art presents, implements, and tests the use of simple single-output classification and regression models for the purpose of art generation. Various machine learning algorithms are trained on collections of computer generated images, artworks from Vincent van Gogh, and artworks from Rembrandt van Rijn. These models are then provided half of an image and asked to complete the missing side. The resulting images are displayed, and we explore implications for computational creativity.