Smelly, dense, and spreaded: The Object Detection for Olfactory References (ODOR) dataset

• Published in: Expert systems with applications 2024-12, Vol.255, p.124576, Article 124576
• Main Authors: Zinnen, Mathias, Madhu, Prathmesh, Leemans, Inger, Bell, Peter, Hussian, Azhar, Tran, Hang, Hürriyetoğlu, Ali, Maier, Andreas, Christlein, Vincent
• Format: Article
• Language: English
• Subjects: Computational humanities; Dataset; Object detection; Olfaction; Small object detection; Transfer learning
• ISSN: 0957-4174
• DOI: 10.1016/j.eswa.2024.124576

Real-world applications of computer vision in the humanities require algorithms to be robust against artistic abstraction, peripheral objects, and subtle differences between fine-grained target classes. Existing datasets provide instance-level annotations on artworks but are generally biased towards the image centre and limited with regard to detailed object classes. The proposed ODOR dataset fills this gap, offering 38,116 object-level annotations across 4712 images, spanning an extensive set of 139 fine-grained categories. Conducting a statistical analysis, we showcase challenging dataset properties, such as a detailed set of categories, dense and overlapping objects, and spatial distribution over the whole image canvas. Furthermore, we provide an extensive baseline analysis for object detection models and highlight the challenging properties of the dataset through a set of secondary studies. Inspiring further research on artwork object detection and broader visual cultural heritage studies, the dataset challenges researchers to explore the intersection of object recognition and smell perception. •Dataset of 4712 artworks annotated with 38116 labelled object instances from 139 categories.•First dataset of smell-related objects in artworks.•Challenging dataset in terms of occlusion, object sizes, and spatial object distribution.•Extensive performance analysis of common object detection algorithms.