Dataset of arthropod flower visits captured via smartphone time-lapse photography

This annotated dataset comprises time-lapse images of arthropods visiting flowers, captured using smartphones placed above selected open flowers. The data collection took place from July to September 2021 in urban green spaces in and around Leipzig and Halle, Germany. Each arthropod in the images is marked with a bounding box, and individuals from the Hymenoptera (excluding ants) and Diptera orders have been identified to the lowest possible taxonomic level. This dataset repository corresponds to the published article: Ștefan V., Stark T., Wurm M., Taubenböck H., Knight T.M. (2025). Successes and limitations of pretrained YOLO detectors applied to unseen time-lapse images for automated pollinator monitoring. Sci Rep 15, 30671 (2025). DOI: https://doi.org/10.1038/s41598-025-16140-z The detailed methodology of data collection is provided in: Ștefan, V., Workman, A., Cobain, J. C., Rakosy, D., & Knight, T. M. (2025). Utilising affordable smartphones and open-source time-lapse photography for pollinator image collection and annotation. Journal of Pollination Ecology, 38, 1-21. DOI: https://doi.org/10.26786/1920-7603(2025)778 This research was funded by the Helmholtz AI initiative (Information & Data Science) Pollination Artificial Intelligence (ZT-I-PF-5-115), led by Prof. Tiffany M. Knight and Prof. Hannes Taubenböck, the Helmholtz Recruitment Initiative of the Helmholtz Association to Tiffany M. Knight, and iDiv (German Research Foundation FZT 118) DATASET LICENSE & USAGE POLICY Copyright Holder: Copyright (C) 2026 Valentin Ştefan, Helmholtz Centre for Environmental Research (UFZ) & German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig. 1. PRIMARY LICENSE This dataset is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0). 2. AI & MACHINE LEARNING RESTRICTIONS The exclusive rights to use this dataset, including any modified or derivative versions, for the purpose of training, fine-tuning, or developing commercial Artificial Intelligence (AI) models are expressly reserved by the Copyright Holder. 3. TDM RESERVATION (EU LAW) Pursuant to Article 4(3) of Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market, the Copyright Holder hereby expressly reserves the use of the works contained in this dataset for text and data mining (TDM) for commercial purposes. 4. SCIENTIFIC & NON-COMMERCIAL USE Use for non-commercial academic research, including the development of open-source, non-commercial scientific models, is permitted and encouraged under the terms of the CC BY-NC-SA 4.0 license, provided appropriate attribution is given.