Alexander Mckeeken

Major: BCB

Faculty Advisor: Marek Borowiec

Project Title:

Can We Automate Bug Identification? A Case Study Using Deep Learning to Identify Photographs of North American Arthropods

Abstract

This project attempts to develop a way in which we can quickly and accurately identify arthropod specimens in a range of taxonomic classifications such as family, genus, and species. This problem is important because identification is an integral step to any downstream biological application such as designing experiments dealing with species-species interactions. However, the way in which we approach identification requires expert knowledge, is often time-consuming, and may be biased or inaccurate. Finding ways to minimize these issues could lead to more efficient and better-quality research in the future. We address this by training a deep learning model that can identify North American arthropod genera from images. Deep learning is a relatively new class of machine learning algorithms that can solve complex problems that are too difficult to program by hand. This approach has skyrocketed in popularity due to advances in both data collection and hardware used for deep learning. We chose to collect our data from two sources, BugGuide.net and iNaturalist.org, because they provide us with a large database of images and complement each other. BugGuide.net offers high-quality, carefully curated images while iNaturalist.org contains more images of common species captured at often lower quality. No existing deep learning identification tool focuses on North American arthropods. iNaturalist’s identification tool “Seek” attempts automated identification of all living organisms at the expense of accuracy. In the future we plan to refine our model make it available as a phone application and a website. We are also planning to integrate support for live video format processing and object detection. These features will allow researchers and the public to use our system more easily and hopefully lead to integration of our technology into larger projects.

Funding: University of Idaho (UI ORED 2019 Jumpstart grant and PI’s startup fund)