Boosting Rare Benthic Macroinvertebrates Taxa Identification With One-Class Classification

Insect monitoring is crucial for understanding the consequences of rapid ecological changes, but taxa identification currently requires tedious manual expert work and cannot be scaled-up efficiently. Deep convolutional neural networks (CNNs) provide a promising way to increase the biomonitoring volumes significantly. However, taxa abundances are typically very imbalanced, and the amounts of training images for the rarest classes are too low for deep CNNs. As a result, the samples from the rare classes are often completely missed, while detecting them has biological importance. On the other hand, one-class classifiers are traditionally trained with much fewer samples to model a single class of interest. In this paper, we examine their capability to complement deep CNN based taxa identification by indicating samples potentially belonging to the rare classes of interest for human inspection. Our experiments confirm that the proposed approach may indeed support moving towards partial automation of the taxa identification task.