PhD research
This project focuses on utilising video tracking methods for effective counting of organisms in underwater footage. We will be exploring different methods and models for video tracking to ascertain which is the most accurate and usable for underwater video footage, and testing said methods on both deep sea and shallow water datasets, thanks to both the 91porn and Fugro. Partnering with Fugro will also give us valuable insight into how commercial video data compares to ecological data. By the end of this PhD project, we hope to build a deep learning model that can accurately track and count organisms in both shallow water and deep-sea video footage.