Monitoring Aotearoa’s native bird species and identifying at-risk populations are critical for their conservation. While high-risk species require detailed population monitoring, which is often intrusive (capturing and fitting radio/GPS transmitters to birds), it is desirable to have an alternative low-cost, passive solution. With many visually cryptic (hard to see) remote species, such as kiwis, acoustic methods involving animal call tracking are ideal. The supervisor’s team has developed a prototype of a passive acoustic monitoring system through the National Science Challenge SfTI projects (https://www.sftichallenge.govt.nz/news/award-winning-science-communications-videos-from-first-time-directors/) funded by the Ministry of Business Innovation and Employment. The system localises and identifies bird species using audio signal processing and machine learning algorithms. While the prototype has proven that the concept is promising, the ultimate goal is enabling the system to operate for weeks or months in remote forests without replacing batteries, which is not a trivial task due to the extensive computational costs associated with the contemporary machine learning and signal processing algorithms used.
This project aims to analyse the computational costs associated with the bird monitoring system to optimise its overall performance. The focus will be on exploring the computational efficiency of various machine learning algorithms used for the identification of bird calls. The investigation will encompass the evaluation and comparison of the computational costs associated with these algorithms. The students taking this project will have opportunities to learn the state-of-the-art technologies in machine learning and signal processing and will contribute to the conservation activities to protect precious birds in Aotearoa. Having experience with Python and Matlab programming would be advantageous.
The supervisor is leading the Communication Acoustics Lab (CAL) at the Acoustics Research Centre. To see how this project is related to the research conducted by the CAL, visit the website from this link: http://cal.auckland.ac.nz.
Undergraduate
As this project requires fundamental knowledge in digital signal processing, at least one of the group members must have passed MECHENG370 Electronics and Signal Processing. Due to the nature of the project, it is availble only to Mechatronics students.
Acoustics Lab (City 422.154, Lab)