Postgraduate research project

Advanced speech enhancement using Transformer networks

Funding
Competition funded View fees and funding
Type of degree
Doctor of Philosophy
Entry requirements
2:1 honours degree View full entry requirements
Faculty graduate school
Faculty of Engineering and Physical Sciences
Closing date

About the project

This PhD will develop new AI for clearer speech, focusing on helping those with hearing loss understand conversations better, especially in noisy places. The goal is to improve hearing aids and assistive tech using cutting-edge Transformer networks.

This PhD project will explore the novel application of Transformer networks for speech enhancement, specifically aimed at improving intelligibility for hearing-impaired individuals. While Transformers have shown success in other speech tasks, their use for targeted intelligibility enhancement remains largely unexplored. This research will delve into developing specialized Transformer architectures optimized for this purpose.

You will:

  • investigate novel Transformer architectures tailored for speech intelligibility enhancement.
  • develop training strategies and data augmentation techniques for robust performance in real-world noisy environments
  • evaluate the effectiveness of the developed models through objective metrics and subjective listening tests with hearing-impaired participants

This research has the potential to significantly improve the quality of life for hearing-impaired individuals by making speech more intelligible in challenging listening situations. Successful outcomes could lead to advancements in hearing aid technology, assistive listening devices, and other applications that benefit from enhanced speech clarity.