Applying Bioacoustics and Artificial Intelligence to study the spatial ecology of endangered species in Borneo

Project Background

Tropical rainforests are some of the most biodiverse habitats on earth, but they are widely impacted by human activities. Therefore, they are home to many endangered species that require careful monitoring of their remaining populations. However, monitoring animals in rainforests is often very challenging because the animals are shy, difficult to see in dense vegetation and because these forests by are typically relatively inaccessible.

As a result, estimates of population densities of even large charismatic species often have a high degree of uncertainty, poor spatial resolution and are difficult to repeat consistently over time.

In recent years, there has been steep growth in the use of Passive Acoustic Monitors (PAM) that record sound continuously. PAMs deployed over large areas and over longer periods of time have great potential to provide consistent data on the amount of vocal activity of specific species over space and time. The main limiting factor in the utility of PAMs currently lies in extracting the necessary information from the vast amounts of acoustic data that they produce.

The effective application of Artificial Intelligence and other aspects of Data Science are therefore essential for moving this field forward. Project Aims and Methods

This project benefits from a large collection of recordings collected by us and project partners in Indonesian Borneo over the last 5-years. The initial focus will be on applying machine learning solutions to White-Bearded Gibbon detection in these recordings and using this to estimate population densities and other aspects of this species' ecology.

Secondly, the aim is to use lessons learned from this process to create and test a general workflow for applying the methodology to a wide range of species. There will be substantial flexibility in the specific direction of the research within these broad aims, depending on the interests of the student and priorities identified by project partners.

Project partners

The student will spend at least three months on placement with Borneo Nature Foundation International, which provides an excellent opportunity to gain experience with all aspects of operations in a conservation NGO. In addition, the student will benefit from opportunities to interact with our active international network of collaborators