This dataset, titled "Digital Humanities Retrieval Evaluation Dataset for Art, Architecture, and Life Sciences," emphasizes semantic retrieval, including ground truth data for both content-based and instance-based image retrieval. Utilized to compare various promising approaches and refine the evaluation method, this dataset serves as a precursor to a more extensive dataset. Additional details can be found in the dataset's related publication. Currently, the dataset is still under development. In its next iteration, the dataset will include options for retrieving image syntax and temporal aspects, in addition to the semantics.