A FAIR annotated set of 60k+ in situ zooplankton images from the Greater North Sea and Greenland

This is the dataset accompanying the paper "A FAIR annotated set of 60k+ in situ zooplankton images from the Greater North Sea and Greenland" that was submitted to Earth System Science Data. This dataset contains i) annotated images of zooplankton and particles of CPICS and ISIIS ii) two CNN classifiers - one for the ISIIS and one for the CPICS - based on this annotated set iii) Python-code to (re)train the CNNs and to run these on new images and iv) code and supporting files to reproduce all figures presented in the manuscript. The annotated dataset contains labeled images of zooplankton, detritus and other particles collected with the ISIIS-DPI and CPICS in the Greater North Sea region and Greenland between 2021 and 2024. Data were annotated following a standardised labeling protocol that is explained in the manuscript and in the file readme.txt, and that was based on the preprint "PlanktoShare: A large (50k+) and FAIR learning set for the Plankton Imager (Pi-10) for the Greater North Sea and NE Atlantic, based on a new flexible classification protocol" by Van Walraven et al. (https://doi.org/10.5194/essd-2026-215) The CNN classifiers and the shared Python-code can be used to classifiy newly collected images of CPICS and ISIIS within the study region, or can be retrained using subsets or aggregations of the annotated data depending on the application at hand. A detailed description of all shared files is in readme.txt. All code needed to run the Python-scripts and notebooks are shared in code.zip. Updated versions of the software will be shared at https://github.com/pieterhovenkamp/AnImageOfContrasts