Penguin Behaviour Classification

This project, titled, An Investigation into the Automatic Behaviour Classification of Penguins, evaluates the combined contribution of pose estimation and behaviour classification as applied to the endangered African penguin.

This upload comprises the tutorial document, GUI, annotation data, and video data facilitating this project. Data investigating one, two, and multiple (three) individuals are provided. The one individual case study considers the following behaviours: braying (b), flapping (flap), preening (preen), resting (rest), standing (stand), and walking (walk). The two individual case study considers the following two behaviours: interaction (int) and no interaction (nint). Finally, the multi-individual case study considers the following two behaviours: excitement (ab) and normal (norm). Combined, these provide a collective overview of the data studied in this project and yield insight into applying machine learning to ecology.