Using AI to clear back logs of camera trap data at National Wildlife Refuges and beyond

In 2022, I became the project manager for a new cooperative agreement with US Fish and Wildlife to help clear a backlog of unsorted camera trap images from National Wildlife Refuges throughout the southwest region. I built a participatory science project using the Zooniverse platform to collect labels for over a dozen common species found at wildlife management areas in New Mexico, Texas, and Oklahoma. While previously we were training site-specific deep learning models, we are experimenting with the new SpeciesNet detector and classifier. Currently, we are also partnering with New Mexico State University and the Turner Reserve to clear a decade-long backlog of camera trap imagery from the Armendaris Ranch in southeastern New Mexico.

As an outgrowth of this work, I have also been assisting on a project to derive distance images using photogrammetry for the purpose of ungulate population density estimation at Sevilleta NWR and San Andres NWR.