Dataset used in training for parts of OntoUML Image Taxonomy Extractor (OITE)
This dataset consists of the 643 OntoUML images from the OntoUML Catalogue, 643 randomly picked class diagrams and 643 randomly picked non-UML images, taken from the dataset of the paper “Multiclass classification of four types of UML diagrams from images using deep learning". In my paper this dataset is referred to as the 3-class dataset.
See: https://github.com/SimeonKaishev/OntoUML_IMG_Converter