A position is available to work on machine learning methods for computer graphics with the starting date in October 2019. The project aims to create images that would appear highly realistic to the human eye, to the point that they are confused with real-world objects. This will be achieved by combining high dynamic range, binocular stereo and multiple focal depths in a single display setup. The goal is to determine what factors are the most relevant for achieving highly realistic look of the displayed scenes. The outcome of the project will help to define requirements for future display technologies, rendering, representation and storage of visual content. The successful candidate is expected to work on highly realistic rendering of 3D scenes using machine learning and image-based rendering methods. Such rendered scenes will be shown on custom-built displays that combine high dynamic range with binocular stereo and focal depth cues.
Candidates should have a 1st class degree in computer science, electronic engineering, mathematics or a closely related discipline, with experience and interest in machine learning and computer graphics. Excellent programming skills are required.
A candidate meeting the required PhD admission criteria will be invited to apply to the PhD programme in Computer Science.
It is desirable that a candidate has experience in computer graphics and/or computer vision. If a candidate has a track record of publications, it should be included in the application as a link to Google Scholar or ORCID profile.
This position is funded through the European Research Council (ERC) Consolidator Grant. If the candidate registers as a full-time PhD student at the University and study for the degree, the candidate will be eligible to pay staff rate tuition fees for the PhD, which for 2019-20 will be £2,619 per annum for three years. The position is open to both UK/EU and non-EU applicants.