We are looking for a PhD candidate interested in ecology, predictive habitat modelling, remotely piloted aircraft systems (RPAS; or drones), and/or machine learning to join the Centre For Integrative Ecology at Deakin University within an exciting research project developing improved monitoring techniques for koala and kangaroo populations and habitat.
Conservation practitioners and land managers require the most robust, efficient, accurate, and cost-effective wildlife monitoring techniques in order to manage species and their diverse habitats across vast Victorian landscapes.
Survey methods that provide evidence to support critical wildlife decisions such as relocation, control or take no action, are often contested, due to questions around the accuracy, cost, difficulty, and time required to undertake reliable surveys. Emerging technologies such as compact remote sensing technology, remotely piloted aircraft systems (RPAS; or Drones), rapid digital image processing and machine learning technologies provide opportunities to improve the efficiency and accuracy of wildlife monitoring and habitat assessments.
This project aims to develop, implement and evaluate a range of these new approaches to monitor wildlife populations and habitat conditions. The project will incorporate conventional field-based surveys and emerging aerial survey techniques (including drones) combined with machine (deep) learning to identify improved monitoring approaches for koalas and kangaroos and their habitats.
The Bushfires and Natural Hazards Cooperative Research Centre-funded project is an exciting opportunity to collaborate and study across disciplines. The project draws on expertise from the Deakin Schools of Life and Environmental Sciences, Information Technology, and Business and Law. The project also involves collaborations across government, including the Victorian Department of Environment, Land, Water and Planning, the Arthur Rylah Institute, and Parks Victoria.
The specific roles of the PhD candidate include:
- Build and test predictive models to characterise vegetation quality for koalas and kangaroos using drone and ground-collected data.
- Assist IT experts in developing continuous scanning machine learning using recently developed heterogeneous data fusion techniques based on fuzzy Choquet integration of deep convolutional neural network classifiers. The candidate will collaborate to produce a novel automated vertebrate detection architecture from multispectral imagery captured from drones for koalas and kangaroos.
- Assist in koala and kangaroo field work in remote areas (e.g., SW Victoria, Otways or South Gippsland for koalas & NW Victoria, Murray-Sunset, Hattah-Kulkyne and Wyperfeld national parks for kangaroos). Field work will include conducting conventional field surveys and supporting drone pilot teams as a spotter.
- Annotate drone-collected video data sets for koalas, kangaroos and habitat, including visual, multispectral and thermal footage of habitat condition and koala/kangaroo detections.
- A relevant undergraduate degree (e.g., degree in IT/computer science or science and/or environmental science) in a relevant area (e.g., GIS and remote sensing, spatial science, machine learning, biodiversity conservation/ecology).
- Class 1 Honours or Masters, also in a relevant area.
Knowledge, skills and experience:
- Demonstrated ability to work independently as well as collaborate and work effectively in a team-based environment. The successful candidate will study in a large and diverse project team involving partners in government departments and research institutions.
- Demonstrated quantitative skills in data analysis.
- Proven scientific writing skills.
Undergraduate, research and/or technical experience (or a demonstrated capacity to develop skills) in the below listed areas would be an advantage:
- Familiarity/experience with machine learning and artificial intelligence.
- Predictive habitat modelling and quantitative spatial ecology/remote sensing.
- Survey/monitoring techniques for large mammals and/or arboreal mammals.
- Qualifications and/or experience with remotely piloted aircraft systems (RPAS; or drones).
- Capacity to undertake efficient ecological field work in remote areas.
The successful candidate will be awarded a 3-year PhD Scholarship (~AU$28,000 p.a. tax free) and will be based at Deakin University’s Burwood campus (Melbourne). Please note: the broader research project has commenced, therefore applicants who are able to start in the first quarter of 2022 will be prioritised. While all applicants will be welcomed, due to COVID-19 and the preferred starting period, domestic applicants may be prioritised.
Interested in applying?
Interested candidates should contact Lachlan Howell (firstname.lastname@example.org), the Principal Research Fellow on this project. The candidate should forward the following documents: 1) An updated CV highlighting their skills, education, relevant work experience and any publications; 2) A cover letter (no more than 3 pages) that specifically addresses each of the selection criteria, including an introductory paragraph outlining their interest in the position.
Closing date: Wednesday, 2nd February 2022 @ 5pm
For more available positions within the CIE please visit our Current Vacancies page.