Aesthetics Without Semantics

TL;DR

This study introduces a new image database with minimal semantic content and a method to generate ugly images, enabling better understanding of aesthetic judgments independent of semantic influence.

Contribution

The paper presents the MSC database of over 10,000 images and a technique for generating aesthetically ugly images, addressing biases in existing datasets and advancing aesthetic research.

Findings

Augmenting datasets with ugly images alters feature-aesthetics relationships.

Bias towards beautiful images in datasets can distort aesthetic analysis.

Including diverse aesthetic values reveals complex feature-judgment interactions.

Abstract

While it is easy for human observers to judge an image as beautiful or ugly, aesthetic decisions result from a combination of entangled perceptual and cognitive (semantic) factors, making the understanding of aesthetic judgements particularly challenging from a scientific point of view. Furthermore, our research shows a prevailing bias in current databases, which include mostly beautiful images, further complicating the study and prediction of aesthetic responses. We address these limitations by creating a database of images with minimal semantic content and devising, and next exploiting, a method to generate images on the ugly side of aesthetic valuations. The resulting Minimum Semantic Content (MSC) database consists of a large and balanced collection of 10,426 images, each evaluated by 100 observers. We next use established image metrics to demonstrate how augmenting an image set biased towards beautiful images with ugly images can modify, or even invert, an observed relationship between image features and aesthetics valuation. Taken together, our study reveals that works in empirical aesthetics attempting to link image content and aesthetic judgements may magnify, underestimate, or simply miss interesting effects due to a limitation of the range of aesthetic values they consider.