Will artificial intelligence revolutionize aerial surveys? A first large-scale semi-automated survey of African wildlife using oblique imagery and deep learning

A semi-automated deep learning approach significantly enhances the efficiency and accuracy of wildlife population estimation in aerial surveys, offering comparable or superior results to traditional methods. The study, conducted in Comoé National Park, Ivory Coast, combined oblique-camera-count (OCC) imagery with a pre-trained deep learning model, HerdNet, to process over 190,000 images. This method reduced human effort by 111 hours while improving detection of small and isolated animals, such as kob and warthog, compared to the rear-seat observer (RSO) approach. Despite challenges like vegetation obstruction, the semi-automated deep learning (SADL)-OCC pipeline demonstrated transformative potential for future wildlife monitoring, particularly with microlight aircraft and drones.

Delplanque, A., Linchant, J., Vincke, X., Lamprey, R., Théau, J., Vermeulen, C., Foucher, S., Ouattara, A., Kouadio, R., & Lejeune, P. (2024). Will artificial intelligence revolutionize aerial surveys? A first large-scale semi-automated survey of African wildlife using oblique imagery and deep learning. Ecological Informatics, 82, 102679. https://doi.org/10.1016/j.ecoinf.2024.102679