ICC++: Explainable feature learning for art history using image compositions. (April 2023)

• Record Type: Journal Article
• Authors: Madhu, Prathmesh; Marquart, Tilman; Kosti, Ronak; Suckow, Dirk; Bell, Peter; Maier, Andreas; Christlein, Vincent
• Is Part Of: Pattern recognition. Volume 136(2023)
• Publication Date: 2023-04
• Subjects: Image/Scene compositions -- Computer vision -- Art history; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2022.109153
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: ICC++ is explainable and content-based image retrieval method. ICC++ is based on compositional patterns for semantic understanding of art historical images. Introducing Web Gallery of Art 500 (WGA500) dataset - to encourage further research in this domain. Rigorous evaluations against traditional methods, deep features & SOTA methods. Our method outperforms SOTA, while has an advantage over deep features in terms of explainability. Abstract: Image compositions are helpful in the study of image structures and assist in discovering the semantics of the underlying scene portrayed across art forms and styles. With the digitization of artworks in recent years, thousands of images of a particular scene or narrative could potentially be linked together. However, manually linking this data with consistent objectiveness can be a highly challenging and time-consuming task. In this work, we present a novel approach called Image Composition Canvas (ICC + + ) to compare and retrieve images having similar compositional elements. ICC + + is an improvement over ICC, specializing in generating low and high-level features (compositional elements) motivated by Max Imdahl's work. To this end, we present a rigorous quantitative and qualitative comparison of our approach with traditional and state-of-the-art (SOTA) methods showing that our proposed method outperforms all of them. In combination with deep features, our method outperforms the best deep learning-based method, opening the research direction for explainable machine learning for digital humanities. We will release the code and the data post-publication.