Towards a Taxonomy Machine. A Training Set of 5.6 Million Arthropod Images

Dirk Steinke; Sujeevan Ratnasingham; Jireh R. A. Agda; Hamza Ait Boutou; Isaiah C. H. Box; M Boyle; D Chan; Corey Feng; SC Lowe; Jaclyn McKeown; Joschka McLeod; A Sanchez; I Smith; Spencer Walker; Catherine Wei; Paul D. N. Hebert

doi:10.3390/data9110122

BioinformaticsDataOct 25, 2024

Peer-reviewed journal article

Towards a Taxonomy Machine. A Training Set of 5.6 Million Arthropod Images

Dirk Steinke iD, Sujeevan Ratnasingham iD, Jireh R. A. Agda iD, Hamza Ait Boutou iD, Isaiah C. H. Box iD, M Boyle, D Chan iD, Corey Feng iD, SC Lowe, Jaclyn McKeown iD, Joschka McLeod iD, A Sanchez, I Smith, Spencer Walker, Catherine Wei iD, Paul D. N. Hebert iD

Abstract

The taxonomic identification of organisms from images is an active research area within the machine learning community. Current algorithms are very effective for object recognition and discrimination, but they require extensive training datasets to generate reliable assignments. This study releases 5.6 million images with representatives from 10 arthropod classes and 26 insect orders. All images were taken using a Keyence VHX-7000 Digital Microscope system with an automatic stage to permit high-resolution (4K) microphotography. Providing phenotypic data for 324,000 species derived from 48 countries, this release represents, by far, the largest dataset of standardized arthropod images. As such, this dataset is well suited for testing the efficacy of machine learning algorithms for identifying specimens to higher taxonomic categories.

PDF DOI