Automatic Digital Classification for Insect Pests: Out of the Lab, and Into the Field

Monitoring the presence and density of insects, typically down to the level of sex, species and lifestage, is critical for optimizing interventions for agriculture, medical, veterinary and ecological applications. This seems like a natural problem for Artificial Intelligence (machine learning). However, the problem is difficult for many reasons; there are tens of thousands of insects that may be of interest, and they may look, sound, behave and even smell similar to each other because they are sibling species or mimics, or because of convergent evolution etc. This talk will show that the machine learning community has made significant progress in automatic digital insect classification, at least in the lab and will further illustrate that many of the remaining barriers to widespread adoption are not computational, but entomological, cultural, regulatory and pragmatic.

About the speaker

Eamonn Keogh, Ph.D. is a distinguished professor of computer science at the University of California-Riverside and a co-founder of FarmSense, a digital insect surveillance company. His work on computational entomology has won many awards including the Bayer Prize for Insect Classification, The National Centre for Palms & Dates International Prize, St. Patrick’s Day Medal (Ireland’s top science prize), Bell Lab's Prize, a Bill and Melinda Gates Grand Challenge and the Vodafone Wireless Innovation award.

Cornell IPM Academic Seminars

This event is part of the Cornell Integrated Pest Management academic seminar series which is designed to increase awareness of new research and techniques that advance Integrated Pest Management (IPM) and its adoption in all types of pest management settings.