Image-Based Recognition of Parasitoid Wasps Using Advanced Neural Networks

Abstract

Hymenoptera have some of the highest diversity and number of individuals among insects. Many of these species potentially play key roles as food sources, pest controllers, and pollinators. However, little is known about their diversity and biology, and about 80% of the species have not been described yet. Classical taxonomy based on morphology is a rather slow process, but DNA barcoding has already brought considerable progress in identification. Innovative methods such as image-based identification and automation can even further speed up the process. We present a proof of concept for image data recognition of a parasitic wasp family, the Diapriidae (Hymenoptera), obtained as part of the GBOL III project. These tiny (1.2 - 4.5 mm) wasps were photographed and identified using DNA barcoding to provide a solid ground truth for training a neural network. Subsequently, three different neural network architectures were trained, evaluated, and optimized. As a result, 11 different classes of diaprids and one class of "other Hymenoptera can be classified with an average accuracy of 96%. Additionally, the sex of the specimen can be classified automatically with an accuracy of > 96%.