Semantic indexing and computational aesthetics: Interactions, bridges and boundaries doctoral symposium

Semantic Indexing and Computational Aesthetics are two closely related fields. For some aspects they are similar, complementary for others, and sometimes completely disjoint. Semantic Indexing is about automatically identifying content in natural images, namely recognizing objects and scenes. Computational Aesthetics provides a set of techniques
to automatically assign a beauty degree to a given image. In our work, we enrich both types of visual analysis by exploring the synergy of those two fields. We investigate the role of Semantic Indexing techniques for Computational Aesthetics Frameworks, and, vice versa, the importance of Aesthetic features for Semantic Indexing prediction. We show the benefits and the limits of this synergy, and propose some improvements in this direction.