T Hain

This fully-funded PhD studentship at the University of Sheffield offers an exciting opportunity to advance research in metamodelling of speech domains. Speech signals are inherently variable and often recorded in challenging environments, containing not only spoken words but also information about speaker identity and conversation style. The project aims to explore how different aspects of speech data are represented in model space, particularly in the context of automatic speech recognition and diarisation. Recent advances in speech processing have seen model topologies converge, driven by a focus on generalisation, yet domain dependence remains a challenge. This research will investigate novel methods for representing model variations and attributing them to specific tasks and attributes. The outcomes will inform training and inference processes, with strategies including new approaches to model distributions and parameter predictions. Experiments will span tasks such as speech classification, recognition, and diarisation across diverse data domains. You will join the Speech and Hearing research group within the School of Computer Science, a department recognised for world-leading research (99% rated in the highest two categories in REF2021). The studentship is affiliated with the UKRI AI Centre for Doctoral Training (CDT) in Speech and Language Technologies (SLT) and their Applications, providing access to a vibrant research community and training opportunities. Funding covers Home or International tuition fees for 3.5 years, an enhanced tax-free stipend (£25,305 for 2026/27), and a research and training support grant (£2,500 per annum pro rata) for research expenses and conference attendance. Applicants should have a strong academic background in computer science, mathematics, or related fields, with experience in speech processing, machine learning, or data analysis desirable. Both Home and International students are eligible, and detailed eligibility and application guidance are available via the application portal. The application deadline is 23:59 on 19 April 2026. Shortlisted candidates will be invited to interview in Sheffield or via videoconference in early- to mid-May. If positions remain unfilled, a rolling review and interview process will be implemented. For informal enquiries, contact [email protected]. Apply online via the UKRI AI Centre for Doctoral Training application portal. For more information, visit the project page or the application guidance link.