An Alchemical Object Detection Dataset for Early Modern Scientific Illustrations

This annotation data is based on an image dataset curated by Ute Frietsch at the HAB as part of a previous project that involved the development of Iconclass classifications for alchemical objects. In "Erschließung alchemiegeschichtlicher Quellen" at the Herzog August Library, a dataset of 1,800 relevant book pages from early modern prints was tagged with keywords (https://alchemie.hab.de/bilder).

To make the data usable for computer vision and specifically object detection, the locations of the objects on the images need to be specified in a usable image annotation format (such as MS COCO or YOLO) so they can be used to train machine learning algorithms. In the context of a 2024 Wolfenbüttel NFDI 4 Memory Fair Data Fellowship, approximately 640 of these pages were classified using the Supervisely platform to provide ground truth data in the annotation formats mentioned above. These make up the contents of this repository along with a README from the related GitHub repository (https://github.com/sarahalang/hab-nfdi4memory-fairDataFellow). The dataset is described in more detail in the data paper, "Fine-Tuning Machine Learning with Historical Data. An Alchemical Object Detection Dataset for Early Modern Scientific Illustrations".