Datasets used in "Stomata Morphology Measurement with interactive Machine Learning: Accuracy, Speed, and Biological Relevance?"

Stomatal imprints of wheat and faba bean, as mentioned in "Stomata Morphology Measurement with interactive Machine Learning: Accuracy, Speed, and Biological Relevance?". A full description of the plant material, the image data, and the interactive machine learning process can be found in the article. Dataset-1 is spring wheat, x10 magnification, with 18 genotypes, grown in a greenhouse. The imprints were collected at stem elongation from three flag leaves per genotype. Dataset-3 is faba bean, x20 magnification, with 6 genotypes, grown outside in the field or in mesocosms. The imprints were collected at leaf development and at flowering from the newest fully developed leaf, with 4 (mesocosm) or 30 (field) replicates per genotype. The validaiton set consits of two plant species, three magnification levels, two imprint methods. Image file names are consisting of: datasetNR_species_ExperimentalEnvironment_ImageMagnification_ImprintMethod_DevevelopmentalStage_Cultivar_LeafSide_LeafPosition_LeafReplicate_ImageReplicate The training set was used to train U-net CNN models with interactive machine learning. All intermediate CNN models are stored.