About the project
This PhD project explores machine learning techniques for affective computing that automates the process of recognising, interpreting and simulating human affects such as facial emotion generation. The current affective computing domain suffers from remarkable expenses and ambiguities on manual annotations.
You will focus on developing novel generative machine learning models to simultaneously synthesise expressions of emotions and learn the emotion label space, by training the machine learning models over either publicly available data or collected data in the group.
The learned label space is expected to be interpretable, complete, and continuous to represent both pure and mixed emotions precisely. Meanwhile, the learned continuous label space enables us to produce more plenary expressions of emotions.
This project will establish a novel direction that machines can automatically learn a more universal emotion labelling system as well as generating more complete expressions of emotions.
You will gain a deep understanding of:
- novel algorithms in computer vision
- machine learning
- affective computing
- generative modelling
You’ll join a team of world leading academics and become part of an extraordinary and vibrant team of postdoctoral researchers and other PhD students. We lead the UKRI Trustworthy Autonomous Systems (TAS) Hub and the UKRI Centre for Doctoral Training in Machine Intelligence for Nano- Electronic Devices and Systems (MINDS). You will be based in the School of Electronics and Computer Science and have access to the University's recently upgraded supercomputer, Idris 5.
These centres and networks will offer you a valuable opportunity to work with researchers and other stakeholders from a wide range of backgrounds. We will support the development of your future career and provide opportunities to network with more than 100 industry partners and universities across the UK. You’ll also have opportunities to expand your experience and knowledge beyond the topic of the PhD project.