Application of Explainable AI methods for taxonomic classification in bird egg collections

This data set is part of a project in cooperation with the University of Warsaw, Poland, dedicated to the development of an AI-based identification method (using Convolutional Neural Networks; CNNs) for bird eggs. Of the target taxon (Western Palearctic species of shrikes, genus Lanius) digital photographs were taken from 1056 eggs (of 175 clutches) from the ornithological collection SNSD (mainly collection W. Makatsch) and size parameters were measured (and if available transcribed from Makatsch’s original catalogue data).