Discovering beautiful attributes for aesthetic image analysis

TL;DR

This paper introduces a method to automatically discover and learn visual attributes from a large aesthetic image dataset, enhancing both accuracy and interpretability in aesthetic image analysis.

Contribution

It proposes a novel approach to automatically learn nameable visual attributes from the AVA dataset, improving aesthetic assessment, tagging, and retrieval.

Findings

Learned attributes improve aesthetic quality prediction.

Attributes enable effective image tagging.

Enhanced image retrieval using learned attributes.

Abstract

Aesthetic image analysis is the study and assessment of the aesthetic properties of images. Current computational approaches to aesthetic image analysis either provide accurate or interpretable results. To obtain both accuracy and interpretability by humans, we advocate the use of learned and nameable visual attributes as mid-level features. For this purpose, we propose to discover and learn the visual appearance of attributes automatically, using a recently introduced database, called AVA, which contains more than 250,000 images together with their aesthetic scores and textual comments given by photography enthusiasts. We provide a detailed analysis of these annotations as well as the context in which they were given. We then describe how these three key components of AVA - images, scores, and comments - can be effectively leveraged to learn visual attributes. Lastly, we show that these learned attributes can be successfully used in three applications: aesthetic quality prediction, image tagging and retrieval.