Vision-Lanuguage Mini Testset (VL-Mini-Test)

A minimal test dataset of 100 images and 12 textual queries for vision-language models, dedicated to the task of text-based image retrieval, and constructed from the following sources: - 50 images from the DocExplore dataset of medieval manuscripts. - 50 images from two manuscripts from Al-Ḥarīrī, <em>Maqāmāt</em>, © Paris, Bibliothèque nationale de France. Département des manuscrits, namely MS arabe 3929 and MS arabe 5847. The research for this work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy – EXC 2176 ‘Understanding Written Artefacts: Material, Interaction and Transmission in Manuscript Cultures', project no. 390893796. The research was conducted within the scope of the Centre for the Study of Manuscript Cultures (CSMC) at Universität Hamburg.