Environmental Futures Seminar - Christopher Burns

Classified maps of coral reef habitats derived from satellite imagery have become essential components for marine science and management. However, current state-of-the-art machine-learning algorithms used to classify these maps lack generalisation capabilities, severely limiting their scalability. To increase the scalability of coral reef habitat mapping from individual reefs to a global scale, machine-learning algorithms capable of generalisation need to be identified. During this exit seminar, I will discuss the results of testing the generalisation capabilities of two machine-learning algorithms in relation to mapping seagrass, reef habitat, and live coral. Additionally, I will discuss the implications of these findings for science and conservation.

Christopher Burns is a PhD candidate supervised by Professor Barbara Bollard, Professor Ajit Narayanan, and Distinguished Professor Sharon Robinson at the School of Earth, Atmospheric and Life Sciences: University of Wollongong, Wollongong, New South Wales, 2522, Australia.