Addressing Racial Bias in Face Recognition Based on Deep Learning Enabled Computer Vision

About the Project

Facial recognition is one of the most academically studied and industrially developed areas within computer vision where we readily find associated applications deployed globally.

This widespread adoption has uncovered significant performance variation across subjects of different racial profiles leading to focused research attention on racial bias within face recognition spanning both current causation and future potential solutions. Building on prior research on this topic at Durham University, where we have addressed topics such as race definition, grouping strategies, and the societal implications of using race or race-related groupings in this domain, we divide the common face recognition processing pipeline into four stages: image acquisition, face localisation, face representation, face verification and identification, and look to study the potential for bias in each and every stage.

There are a wide range of potential research directions, that could form the basis for a specific PhD project in this area:

• identifying and addressing racial bias within the processing pipeline on consumer-level hardware
• use of generative (deep learning) AI approaches as a bias mitigation strategy
• identifying and addressing potential racial bias within both large-scale vision-language (VLM/LLM) and foundational AI models
• exploration of disentangled feature spaces for face generation/adaptation
• face-specific dataset balancing and de-biasing approaches
• applications in face-based tracking across multi-modalities

Supporting Resources

The Department of Computer Science at Durham University hosts well-equipped labs with on-site capabilities comprising vehicle-mounted sensors (LiDAR, radar, camera, GPS/IMU), drone operations, on/off-road robotics, high-precision geo-localisation, BCI/EEG bio-signal data collection, robotic arm manipulation, X-ray security imaging, virtual reality and on-demand wide-area surveillance video feeds.

In addition Durham University hosts the UK regional supercomputer, Bede (128 NVIDIA V100 GPUs) which complements our departmental NVIDIA CUDA Compute Cluster (80+ GPUs up to NVIDIA A100) to cater for the increasing GPU compute demands of modern AI-driven research projects.

The department itself is based in the newly built Mathematical and Computer Sciences building - a £42 million, 9,160m2 state of the art teaching, learning and research facility, located on the University's Upper Mountjoy Campus co-located with the Department of Mathematics. All PhD students are allocated desk and/or lab space to support their research work and PhD student research projects are additionally supported by an annual equipment/travel allowance spanning the duration of the PhD study period.

Supervision

You will be supervised by Prof. Toby Breckon (https://breckon.org/toby/), Professor of Computer Vision and Image Processing in both the Department of Computer Science and Department of Engineering, and Head of Visual Computing (Computer Science) at Durham University in collaboration with one or more staff from the VIViD research group. The work of Prof. Breckon's research team relates to all aspects of computer vision and robotic sensing – the automatic understanding of images by computer as an aspect of artificial intelligence using deep learning (i.e. "visual AI").

Within this domain his team specializes in several industry-facing problem domains spanning X-ray image understanding, automotive vision (autonomous vehicles), visual surveillance, robotic sensing and general topics in object detection/classification. This has resulted in over £19+ million of research income (to 2026+), collaboration with 40+ government and industry partners, over 200 research publications and supported the development of AI software start-up COSMONiO by former team members (acquired by Intel, 2020).

During the PhD study, you will receive extensive training and research guidance via regular one-to-one supervision meetings with Prof. Breckon, in addition to weekly research team meetings, to allow you to both develop your research potential and consolidate your technical skill-base. Prof. Breckons’s research team has a diverse and collaborative culture that aims to facilitate the realisation of each individual's abilities against a range of available research opportunities. Previous PhDs from the team havepublished their research in a range of prestigious venues (e.g. CVPR, ICCV, ECCV, BMVC, WACV) and graduated to a range of careers in the AI industry.

Durham University

Durham University is a world top-100 university (QS), a European top-50 university (QS), ranked the 3rd in the UK (Times, 2026) and a member of the research-intensive Russell Group of UK universities, focussing on research excellence delivered by world-leading academics. Durham is the third oldest university in England, following Oxford and Cambridge, founded in 1832 and situated in the ancient City of Durham, a relatively small city by international standards (pop. ~50,000) situated in the North East of England with long-established historical roots (including an on-campus UNESCO world heritage site comprising Durham Cathedral and Castle). Today the university offers a strong and unique collegiate student experience, with each student being a member of one of the university's 17 constituent colleges, with a relatively low cost of living compared to other parts of the UK.

Entry Requirements

• First class honours (or high 2.1 at undergraduate or equivalent Masters) degree in an Engineering, Physics, Maths or Computer Science based subject (internationally: GPA 3.3+ or equiv).
• Strong understanding of computing/engineering applications and mathematical problem solving.
• Knowledge of modern programming languages (ideally including one of Python, C++/C or Rust).
• Excellent written and spoken communication skills in English, meet the Durham University English language requirements (https://www.dur.ac.uk/study/international/entry-requirements/english-language-requirements/).
• Prior experience in research, industry and/or AI based research topics is beneficial but not essential.

How to Apply

Please send an email with your CV, degree transcripts and any supporting documents to Professor Toby Breckon toby.breckon@durham.ac.uk for an initial pre-application discussion.

Funding Notes

This is primarily a self-funded PhD position and applications are welcome all year round with flexibility on start date.

Whilst Durham University provides fully-funded, highly competitive scholarships for exceptional PhD candidates which open in October each year (including DDS and CSC), the availability of these scholarships is limited and they are awarded on a merit basis each year to our top applicants across all subjects/faculties. For this project, we are happy to consider self-funded students with their own financial support or scholarships from their home nation or another source.