Authors

Discussion

Recently, there has been increased interest in the effect of invasive aphids on other insect and animal species. In order to gain insights on large-scale changes and ecosystem-level impacts, the study of aphids requires the tracking of population trends. The Aphidoidea superfamily is diverse and contains many species. To gain high-level, specific insights into the separate species, populations, and individuals, we propose using deep learning-based computer vision to identify the insects and differentiate between species. However, as with machine learning applications in various fields, potentially the most significant barrier to entry is the curation of an applicable, sufficiently "big" dataset. The imagery can be collected and labeled through crowdsourcing. In this work, we discuss the potential impact and importance of a large scale crowdsourcing effort to obtain a dataset fit for this purpose. We further call upon the community at the burgeoning nexus of entomology and machine learning for collaboration in this area and provide further resources for ideas.