My research intersects deep learning, computer vision, and behavioural ecology, with a particular focus on developing scalable methods for identifying and tracking individual animals in complex and wild marine environments. I am especially interested in uncovering individual-specific signatures across multimodal wildlife data, including visual appearance, vocalisations, movement patterns, and locomotion dynamics, and learning representations that can reliably distinguish individuals across species and environments.
More broadly, I am interested in generalising machine learning methods to complex, data-scarce natural environments, where conventional approaches often struggle with limited annotations, noisy observations, and changing conditions. My goal is to develop robust, data-efficient multimodal models that not only enable non-invasive wildlife monitoring at scale, but also unlock new insights into individual behaviour, movement ecology, and animal–environment interactions.
