In the past 5 years, impressive advances have been made in detecting, recognising and analysing the behaviour of people. Animals, however, have received much less attention despite the importance of these applications in environment conservation, endangered species protection and biosecurity. The goal of this project is to develop deep learning algorithms for the detection, recognition, and the analysis of the behaviour of wild animals in their natural habitat from images captured using standard video cameras. Results of this work will then be used in the management of critically endangered animals such as Gilbert’s Potoroo.
A suitable candidate for this project should have a good expertise in computer vision, image processing, and machine learning. The candidate should also have excellent hands-on programming skills using languages such as Matlab, Python, C and C++.