Using Deep Learning to Unlock Decades of Marine Biodiversity Data and plan marine recovering | DTO-BIOFLOW
A new study published in Ecology and Evolution shows how deep learning can turn decades of archived observations into knowledge about how protection measures and a warming ocean have reshaped life on the seafloor. The research, led by Matthias Obst, marine ecologist at the University of Gothenburg, WP4 leader and coordinator of the Demonstrator Use Cases in DTO-BioFlow, used an AI model trained to recognise 17 benthic species.