StyleBabel: Artistic Style Tagging and Captioning
This work addresses the need for fine-grained artistic style analysis in digital art, benefiting researchers and applications in art and design, though it is incremental as it builds on existing methods like ALADIN.
The paper tackles the problem of describing artistic style in digital artworks by introducing StyleBabel, a dataset of captions and tags for over 135K artworks collected via a participatory method from art experts, and demonstrates its use in tasks like style retrieval, achieving state-of-the-art accuracy.
We present StyleBabel, a unique open access dataset of natural language captions and free-form tags describing the artistic style of over 135K digital artworks, collected via a novel participatory method from experts studying at specialist art and design schools. StyleBabel was collected via an iterative method, inspired by `Grounded Theory': a qualitative approach that enables annotation while co-evolving a shared language for fine-grained artistic style attribute description. We demonstrate several downstream tasks for StyleBabel, adapting the recent ALADIN architecture for fine-grained style similarity, to train cross-modal embeddings for: 1) free-form tag generation; 2) natural language description of artistic style; 3) fine-grained text search of style. To do so, we extend ALADIN with recent advances in Visual Transformer (ViT) and cross-modal representation learning, achieving a state of the art accuracy in fine-grained style retrieval.