"PhD in AI for the Rare and the Uncertain: Improving Camera Trap Species Detection"

University of Exeter - Department of Computer Science

About the Partnership

This project is one of a number that are in competition for funding from the NERC Great Western Four+ Doctoral Training Partnership (GW4+ DTP). The GW4+ DTP consists of the Great Western Four alliance of the University of Bath, University of Bristol, Cardiff University and the University of Exeter plus five Research Organisation partners: British Antarctic Survey, British Geological Survey, Centre for Ecology and Hydrology, the Natural History Museum and Plymouth Marine Laboratory. The partnership aims to provide a broad training in earth and environmental sciences, designed to train tomorrow’s leaders in earth and environmental science. For further details about the programme please see http://nercgw4plus.ac.uk/

For eligible successful applicants, the studentships comprises:

• An stipend for 3.5 years (currently £20,780 p.a. for 2026/27) in line with UK Research and Innovation rates
• Payment of university tuition fees
• The budget for project costs is £9,000 which can be used for computer, lab, and fieldwork costs necessary for you to conduct your research.
• There is also a conference budget of £2,000 and individual Training Budget of £1,000 for specialist training

Project Aims and Methods

Monitoring biodiversity at scale is vital for meeting global conservation targets, yet manual field surveys are too expensive and time consuming. Automated systems like camera traps now generate millions of images of wildlife worldwide, offering unprecedented insights into ecosystems. However, analysing these datasets is a major challenge: AI systems excel at detecting common species but often fail for rare or threatened animals (the “long-tail” problem). Moreover, current models give little indication of when they might be wrong, restricting their use in conservation. This PhD will tackle both issues by developing AI methods that improve recognition of rare species while providing reliable measures of uncertainty. Using state-of-the-art computer vision approaches — vision transformers, self-supervised learning, and few/zero-shot techniques — the student will adapt models to ecological data. Bayesian deep learning and ensemble methods will be explored for trustworthy uncertainty estimation.

The student will benefit from joint supervision between Exeter and Bristol, gaining complementary expertise in AI, computer vision, and ecology. Exeter provides strengths in AI and computer vision, biodiversity monitoring and conservation applications, while Bristol offers advanced training in machine learning, spatiotemporal modelling and AI applications to animal behaviour. Together, they provide computational resources, networks, and training for impactful conservation AI.

Useful recruitment links:

For information relating to the research project please contact the lead Supervisor via: s.rowlands@exeter.ac.uk

Funding Details

For eligible students the studentship will cover home tuition fees plus an annual tax-free stipend.